BULLETIN DE L’INSTITUT ROYAL DES SCIENCES NATURELLES DE BELGIQUE

BULLETIN VAN HET KONINKLIJK BELGISCH INSTITUUT VOOR NATUURWETENSCHAPPEN

ENTOMOLOGIE VOL. 77

BRUXELLES 2007 BRUSSEL Rédacteur en chef - Hoofdredacteur - Editor: Léon BAERT

Redactiesecretarissen - Secrétaires de rédaction - Associate editors Patrick GROOTAERT George WAUTHY

Redactiecomité - Comité de rédaction - Editorial board: Camille PISANI Boudewijn GODDEERIS Jackie VAN GOETHEM

Internationaal comité - Comité international - Consulting editors: Rienk DE JONG (Leiden, The Netherlands) Desmond KIME (Brussels, Belgium) Henk MEUFFELS (Maastricht, The Netherlands) John MURPHY (Hampton, United kingdom) Norman I. PLATNICK (New York, U.S.A.) Michael SCHMITT (Bonn, Germany) Dan BICKEL (Sidney, Australia) Martin BAEHR (München, Germany)

BULLETIN DE L’INSTITUT ROYAL DES SCIENCES NATURELLES DE BELGIQUE ENTOMOLOGIE

BULLETIN VAN HET KONINKLIJK BELGISCH INSTITUUT VOOR NATUURWETENSCHAPPEN ENTOMOLOGIE

Vol. 77 - 2007 ISSN 0374-6232

Publié, verschenen, published: 20.XII.2007

© Edition de © Uitgave van het l’Institut Royal des Sciences Naturelles Koninklijk Belgisch Instituut voor de Belgique Natuurwetenschappen Rue Vautier 29 Vautierstraat 29 B-1000 Bruxelles, Belgique B-1000 Brussel, België CONTENTS

AHRENS Dirk, Revision der Serica nigroguttata 5 LIN Meiying, DRUMONT Alain & YANG Xingke, 147 BRENSKE, 1897 - Gruppe (Coleoptera, The genus Autocrates THOMSON in China: Scarabaeidae) occurrence and geographical distribution of species (Coleoptera: Trictenotomidae) BASSET Yves, et al., IBISCA-Panama, a large- 39 scale study of beta-diversity and LOURENÇO Wilson & QI Jian, Scorpions from the 157 vertical stratification in a lowland rainforest: rainforest canopy of New Guinea and description rationale, study sites and field protocols of a new subspecies of Lychas C.J. KOCH, 1845 (Scorpiones: Buthidae) CONSTANT Jérôme, Revision of the Eurybra- 71 chidae (X). The Oriental genus Chalia WALKER, MOHAMEDSAID Mohamed & CONSTANT 163 1858 (Hemiptera, Fulogoromorpha) Jérôme, Chrysomelid of the subfamily Galerucinae from Thailand and Cambodia in CONSTANT Jérôme, Revision of the Eurybra- 87 the collections of the Royal Belgian Institute of chidae (XI). The Afrotropical genus Mesonytis Natural Sciences (Coleoptera: Chrysomelidae) SCHMIDT, 1908 (Hemiptera, Fulogoromorpha, Eurybrachidae) RYVKIN Alexandr, A brief review of Lathrobium 179 species of the sibiricum-group (Insecta: CONSTANT Jérôme, Note on the coprophily and 107 Coleoptera: Staphylinidae: Paederinae) necrophily Hemiptera Heteroptera TIAN Ming-yi & DEUVE Thierry, Designations 235 FAUCHEUX Michel J., Sensilla and cuticular 113 of the lectotypes for Butes’ species of the genus structures in the marval caudal appendages of Orthogonius MACLEAY (Coleoptera: Caraboidea: Erythromma lindenii (SÉLYS, 1840) (Odonata: Orthogoniini), with descriptions of four new Zygoptera: Coenagrionidae) species FAUCHEUX Michel J., Multiporous and aporous 121 ZHANG Lili, YANG Ding & GROOTAERT Patrick, 243 sensilla on the larval antennae of the relict New Dolichopodidae (Diptera: Dolichopodidae) dragonfly Epiophlebia superstes (SÉLYS, 1889) from Pulau Tioman, Malaysia, with description (Odonata: Anisozygoptera: Epiophlebiidae) of four new species KARISCH Timm, Eine neue Hispophora-Art aus 129 Kenia (Lepidoptera: Geometridae: Ennominae)

KIRK-SPRIGGS Ashley & FREIDBERG Amnon, The 133 Palaearctic species of Curtonotidae (Diptera: Schizophora), with special reference to the fauna of Israel BULLETIN DE L’INSTITUT ROYAL DES SCIENCES NATURELLES DE BELGIQUE ENTOMOLOGIE, 77: 39-69, 2007 BULLETIN VAN HET KONINKLIJK BELGISCH INSTITUUT VOOR NATUURWETENSCHAPPEN ENTOMOLOGIE, 77: 39-69, 2007

IBISCA-Panama, a large-scale study of arthropod beta-diversity and vertical stratification in a lowland rainforest: rationale, study sites and field protocols by Yves BASSET1, Bruno CORBARA2, Héctor BARRIOS3, Philippe CUÉNOUD4, Maurice LEPONCE5, Henri-Pierre ABERLENC6, Johannes BAIL7, Darren BITO8, Jonathan R. BRIDLE9, Gabriela CASTAÑO- MENESES10, Lukas CIZEK11, Aydee CORNEJO12, Gianfranco CURLETTI13, Jacques H. C. DELABIE14, Alain DEJEAN15, Raphael K. DIDHAM16, Marc DUFRÊNE17, Laura L. FAGAN18, Andreas FLOREN19, Dawn M. FRAME20, Francis HALLÉ20, Olivier J. HARDY21, Andrés HERNANDEZ1, Roger L. KITCHING8, Thomas M. LEWINSOHN22, Owen T. LEWIS23, Markus MANUMBOR24, Enrique MEDIANERO25, Olivier MISSA26, Andrew W. MITCHELL27, Martin MOGIA24, Vojtech NOVOTNY11, 24, Frode ØDEGAARD28, Evandro Gama de OLIVEIRA29, Jérôme ORIVEL15, Claire M. P. OZANNE30, Olivier PASCAL31, Sara PINZÓN32, Mathieu RAPP12, Sérvio P. RIBEIRO33, Yves ROISIN21, Tomas ROSLIN34, David W. ROUBIK1, Mirna SAMANIEGO1, Jürgen SCHMIDL7, Line L. SØRENSEN35, Alexey TISHECHKIN36, Christian VAN OSSELAER37 & Neville N. WINCHESTER38

Abstract Key words: biodiversity, canopy, Panama, soil, tropical rainforests

IBISCA-Panama (“Investigating the BIodiversity of Soil and Canopy ”, Panama module) represents a large-scale research initiative to quantify the spatial distribution of arthropod Introduction biodiversity in a Neotropical forest, using a combination of (1) international collaboration, (2) a set of common research questions, Without doubt, one of the fundamental questions in and (3) an integrated experimental design. Here, we present the rationale of the programme, describe the study sites, and outline biology is “how many species are there on Earth”? field protocols. In the San Lorenzo Protected Area of Panama, Sadly, at the beginning of the 21st century, when most twelve 20 x 20 m sites, all less than 2 km apart, were surveyed of the diverse ecosystems of our planet are increasingly for plants and arthropods, from the ground to the upper canopy. threatened by widespread land-use modification, global Access to the canopy and its fauna was facilitated by fogging, climate change, pollution, biological invasions and single-rope techniques and a variety of devices such as a canopy crane, the “SolVin-Bretzel” canopy raft, the canopy bubble and over-harvesting (SALA et al., 2000), there is still no Ikos. IBISCA-Panama represented the first attempt to combine satisfactory answer to this basic question. A related, these complementary techniques of canopy access in a large-scale equally vital, but perhaps methodologically more investigation. Such techniques provided spatial replication during tractable investigation is to assess where the greater initial field work performed in September-October 2003. Temporal replication across seasons consisted of subsequent field work of part of this biodiversity is located. Understanding the varying intensity during dry, early wet and late wet periods in 2004. distribution and ecology of species is also crucial to study Arthropods were surveyed using 14 different protocols targeting the relationship between biodiversity and ecosystem the soil, litter, understorey, mid-canopy and upper canopy habitats. function. By far the greatest fraction of terrestrial These protocols included: WINKLER sifting; BERLESE-TULLGREN; diversity is made up of arthropods. In recent hand-collecting of galls and social ; fogging; beating; wood- rearing; baits; and various types of traps such as pitfall, small and years, there has been considerable debate as to whether large flight-interception, sticky, light, and Malaise traps. Currently, or not most of arthropod biodiversity occurs either in analyses of arthropod distribution in this forest concentrate on a the soil or in the canopy of tropical rainforests (ERWIN, set of 63 focal taxa representing different phylogenies and life- 1982; STORK, 1988; MAY, 1990). This issue lies at the histories. IBISCA-Panama may be considered as a model for large- scale research programmes targeting invertebrate biodiversity. Its heart of wildly varying, and hence hugely contentious, collaborative modus operandi can be applied to answer a variety global estimates of arthropod species richness derived of pressing ecological questions related to forest biodiversity, as from surveys of arboreal arthropods on particular host- evidenced by the recent development of further IBISCA programmes trees (ERWIN, 1982; MAY, 1990, ØDEGAARD, 2000). in other parts of the world. ERWIN (1982) contended that the canopy fauna was the most species-rich. Later, other authors championed the 40 Yves BASSET et al. opposing point of view, that the soil fauna was actually of the studies indicated earlier have sampled canopy more species-rich (HAMMOND, 1990, 1995; ANDRÉ et arthropods in situ with sufficient replication. Within al., 1992; HAMMOND et al., 1997; WALTER et al., 1998). the emerging discipline of canopy biology (OZANNE As yet, however, no data have been collected that are et al., 2003), the most challenging and rewarding extensive enough to test this contention in a convincing scientific advances are likely to be made by studying manner, due principally to a lack of spatial replication the abundance, distribution and functional interactions and a restricted taxonomic focus. among organisms in situ, within relatively undisturbed There have been three previous attempts at a small- rainforest canopies (e.g. LOWMAN & NADKARNI, 1995; scale, restricted comparison of arthropod species STORK et al., 1997; LINSENMAIR et al., 2001; BASSET et richness in the soil-litter layer versus the canopy layer al., 2003a, 2003b). Consequently, biodiversity studies of a tropical rainforest that are worth mentioning. In the (along with studies of atmospheric processes at the Brazilian Amazon, ADIS & SCHUBART (1984) compared canopy interface) figure prominently on the global the fauna of different vertical strata with several research agenda for canopy biology (OZANNE et al., techniques, but they focused primarily on estimates of 2003; DIDHAM & FAGAN, 2004). Although access in the abundance and biomass of arthropod taxa, as opposed situ remains the main limitation, the range of available to species diversity. Similarly, in a 1-ha sampling plot in techniques to reach tree crowns is expanding (review Seram, STORK (1988, 1996) compared the abundance and in MITCHELL et al., 2002) and now allows a freedom of biomass of arthropods extracted from the soil and litter spatial replication in the canopy that was previously using TULLGREN funnels, with those collected from the unheard of. forest canopy using insecticide knockdown (‘fogging’). Second, it is difficult to contrast ground and canopy Both studies focused on abundance and biomass, not arthropod faunas in a directly comparable way, since they on species richness and there was basically no spatial are frequently (and sometimes unavoidably) sampled replication of sites. However, both studies indicated that using different methods. For example, extremely high most of the abundance and biomass of arthropods was densities of springtails have been recorded in canopy concentrated in the soil, not in the canopy. habitats of certain tropical dry forests in Mexico By contrast, the only study that has previously (PALACIOS-VARGAS et al., 1998), but how do these compared the diversity of ground versus canopy densities compare with springtail densities in the soil arthropod communities on a relatively large scale and litter? Habitat structure is very different and it might (accounting to some extent for seasonal variation) was even be difficult to employ the same sampling method performed in Sulawesi (HAMMOND, 1990; HAMMOND in a standardized fashion, making direct comparison of et al., 1997). This intensive sampling programme, ‘sample size’ challenging. Two possible strategies to utilizing multiple sampling methods, suggested that address this problem may be to standardize either to the the soil-litter fauna was indeed richer than the canopy number of individuals collected (species accumulation fauna, with perhaps as many as 70-80 % of species or rarefaction techniques), or to the volume of substrate restricted to the soil-litter habitat (HAMMOND et al., or habitat sampled (or to the number of habitat units 1997). Unfortunately, the study was restricted to beetles sampled). For the former strategy, see for example and there was little replication from which to allow a LINDO & WINCHESTER (2006), comparing mite densities convincing extrapolation of these trends to tropical in ground and suspended soils of temperate rainforests. forests in general (see also STORK & GRIMBACHER, 2006, Third, assessing the relative diversity of soil versus for a study targeting beetles with a single collecting canopy arthropod communities evidently also depends method in Australia). on patterns of beta-diversity. Because of the relatively These pioneering studies have paved the way for high specialization of herbivores on particular more rigorous and intensive studies quantifying the host-tree species (NOVOTNY et al., 2002) and because of spatial distribution of biodiversity by pointing out their associated specific predators and parasitoids, faunal several impediments to clear interpretation of sampling turnover may be rather high in the canopy as compared data, and by highlighting possible strategies to to that in the soil. Thus, it may be inappropriate to circumvent them. First, although sampling arthropods compare the diversity of equivalent projected areas in the soil-litter layer is far from easy (ANDRÉ et al., of canopy and soil, and extrapolate these to predicted 2002), at least spatial replication and methodological estimates of global arthropod diversity. Monodominant ‘repeatability’ within this habitat are somewhat easier stands aside, the beta-diversity of canopy communities to achieve than in the canopy, where access and spatial may be higher than that of soil communities in tropical patchiness of habitats is problematic. In fact, none rainforests. For example, the beta-diversity (and “host IBISCA - Panama 41 specificity”) of soil mites is very low in Australia (OSLER programmes necessary to address the aforementioned & BEATTIE, 2001). In temperate rainforests in Canada, problems are likely to overwhelm investigators and FAGAN et al. (2006) used experimental litter bags to saturate taxonomic experts with the sheer amount of control for microhabitat structure and resource quality, material collected. Aside from the monumental logistical and found some evidence that while ground and canopy difficulties in handling the large numbers of specimens mite assemblages were similar in total biodiversity, collected, the continuing crisis in biosystematics it appeared that local mite richness (alpha diversity) (MILLER, 2000) makes the identification of described was higher on the ground, whereas species turnover species, and the description of new species, collected between sites (beta diversity) was higher in the canopy. in the soil or canopy an increasingly difficult task. For Note, however, that at the appropriate (larger) spatial example, it took two weeks of field work for STORK scale a correlation between below-ground and above- (1991) to fog 15 Bornean trees, but more than a decade ground biodiversity may exist (HOOPER et al., 2000). In for him and his numerous collaborators to sort the particular, plant diversity, because of the production of collected material to morphospecies (unnamed species diverse root exudates, can lead to increased diversity of diagnosed by standard taxonomic procedures), without mutualistic soil microflora, which represents the first accounting for mites. Thus, one has to be extremely link in a cascade of interactions resulting in increased cautious about the expectations of mass collecting diversity of other soil (LAVELLE et al., 1995). programmes and it is not realistic to expect that all the These problems and interactions can be minimised material collected will be studied in the foreseeable and explored, respectively, by considering sampling future. However, a careful selection of focal taxa (i.e., protocols based on spatial replication among soil and a multi-taxa approach), including different orders with canopy habitats at different sites. differing life-history strategies, and a matching of taxa Fourth, microarthropods such as mites are often with motivated and committed investigators, may allow dominant but underestimated in arboreal habitats a robust test of whether spatial distribution patterns of (WALTER & BEHAN-PELLETIER, 1999), whereas they are diverse focal taxa converge or not. Subsidiary funding relatively well sampled in soil and litter. Since Acari may also allow adequate storage of material residues are diverse and numerically dominant in rainforest that may be preserved for future studies. soils (STORK, 1988), comparison between the faunas Despite these obvious impediments, we believe of ground and canopy must ensure that mites have that quantifying the spatial distribution of biodiversity been well sampled in the latter. This, ideally, requires in tropical rainforests and testing hypotheses about similar sampling procedures and samples obtained in the origin and maintenance of this biodiversity are of situ in the canopy. However, this serves to illustrate the fundamentally greater interest than the rather esoteric wider problem of the choice of focal taxa, since many goal of knowing the exact number of species on Earth. groups of tropical arthropods have different ecological The distribution of biodiversity within forests and the requirements and may be expected to follow different functional interactions between ground and canopy distributional patterns (LAWTON et al., 1998). Hence, arthropods are also of central importance to a deeper a multi-taxa approach is likely to be more powerful in understanding of ecosystem dynamics and estimating drawing generalizations about the spatial distribution of levels of redundancy in ecosystem function, and, hence, biodiversity, than concentrating on a single taxon or a are a vital key to forest management for sustainable group of related taxa (see below). use. However, wresting high quality data from tropical Fifth, faunal comparisons rely on the taxonomic forests is a slow and painstaking business, taking study of adult specimens, and immatures are rarely (optimistically) decades of careful study if carried out taken into account in biodiversity assessment, be they by traditional, small research groups (BASSET et al., spiders or beetles. In these taxa, immatures may often 2003a). It is this apparent dilemma – the urgent need develop near or in the soil, and then move up into the for high quality results contrasted with the essentially canopy as adults, and then feed and disperse from there long-term nature of obtaining such results – which led (HAMMOND, 1990; BASSET & SAMUELSON, 1996). Faunal to the development of the IBISCA-Panama programme. comparisons are complicated in this case, but can be Here, we detail the goals of IBISCA-Panama, provide improved by evaluating differences among seasonal descriptions of the study sites and techniques of canopy samples with emergence traps collecting the fauna access, and outline the arthropod sampling protocols that hatches and moves up from the soil (POKON et al., aimed at rigorously quantifying, for the first time, the 2005). spatial distribution of biodiversity in a lowland tropical Finally, the massive scale of the sampling rainforest. This paper is primarily intended to serve as 42 Yves BASSET et al. reference framework for future collective and individual gradients, rather than simply within arbitrary spatial contributions related to IBISCA-Panama. categories. The most important of these environmental variables include: differences in floristic composition, tree basal area, or spatial heterogeneity in vegetation 1. Aims of IBISCA-Panama structure between sites (for spatial turnover among sites); light, canopy openness or leaf area index (for IBISCA-Panama is a research programme that aims vertical stratification); and rainfall, temperature or to quantify the degree of beta-diversity and vertical tree phenology (for temporal variation). We attempt stratification of arthropods in a lowland tropical to predict arthropod distribution patterns from this set rainforest. The programme is based on a combination of variables measured at each site. Progress towards of (1) international collaboration including ecologists, these aims has been updated periodically, as far as is taxonomists, students and parataxonomists; (2) a set possible, in published progress reports from IBISCA- of common research questions; and (3) an integrated Panama (ROSLIN, 2003; BRADBURY, 2003; DIDHAM & experimental design. The large-scale approach is FAGAN, 2003; LONGINO, 2004; SPRINGATE & BASSET, unusual in tropical biodiversity studies and includes 2004; SCHMIDL & CORBARA, 2005; PENNISI, 2005; complementary techniques of canopy access, diverse CORBARA et al., 2006) and news bulletins are available arthropod sampling protocols, a large number of focal at www.ibisca.net. taxa (a multi-taxa approach), substantial spatial and temporal replication, large numbers of experts working 2. Study area and study sites simultaneously in the field, as well as large numbers of 2.1. Study area (Plate 1) taxonomic specialists studying the material collected and analyzing the associated distribution data. As such, The field sampling component of IBISCA-Panama took IBISCA-Panama is an original attempt to overcome place in the San Lorenzo Protected Area (SLPA, Colón the severe impediments on measuring the distribution Province, Republic of Panama), which is situated on of tropical biodiversity, mentioned above. Field work the Atlantic side of the Isthmus of Panama, between was performed at 12 sites in a Panamanian rainforest, Lake Gatun and the Caribbean Sea, near sea level. using 14 arthropod sampling protocols. Many of those This location was chosen because of (a) the presence methods allowed direct, comparative assessment of of the Smithsonian Tropical Research Institute (STRI) arthropod vertical stratification. canopy crane, which facilitated access to the forest The key questions targeted by IBISCA-Panama canopy, in a track of little disturbed, though accessible are: (1) what is the relative contribution of vertical forest; (b) the proximity of a STRI field station with stratification, seasonality and degree of beta-diversity essential laboratory facilities (Barro Colorado Island); to the distribution of arthropod biodiversity in a closed- and (c) its placement in the middle of the Mesoamerican canopy tropical rainforest? and (2) how do life history Biological Corridor “hotspot”, which is known to traits of species, such as host specificity or feeding harbour relatively high levels of biodiversity (WEAVER guild, influence the spatial and temporal partitioning & BAUER, 2004). of arthropod biodiversity in a closed-canopy tropical This lowland wet forest is situated on a geological rainforest? To this end, one leading approach to consider formation known as the Chagres sandstone, dating from is to partition diversity into its spatial and temporal the late Miocene/early Pliocene (PYKE et al., 2001). components (VEECH et al., 2002): total diversity consists This location averages 3139 mm annual rainfall and of alpha diversity (within-sample units), horizontal an annual average air temperature of 26.0ºC (1998- beta-diversity, vertical beta-diversity and seasonal 2002 data). The climate is wet all year round, with a beta-diversity. The interaction between horizontal somewhat drier season between January and mid-April and vertical beta-diversity is of special interest. For (average length of dry season = 125 days: CHAVE et al., the purposes of defining the arthropod entities among 2004). Rainfall and temperature data during the field which to partition components of diversity, we consider study periods (from September 2003 to November three major spatio-temporal ‘axes’ of interest: (a) spatial 2004) are summarized in Fig. 1. The forest is evergreen, turnover among sites, (b) vertical stratification (i.e., with less than 3 % loss in canopy cover by the end of the vertical turnover), and (c) temporal variation among dry season (CONDIT et al., 2000, 2004). The SLPA has repeated sampling intervals through time. Our aims are been mostly free of severe disturbance for the past 150 to contrast these biotic gradients in relation to potential years except for a few isolated spots. As an indication ecological variables that might be driving the spatial of this, individuals of slow growing tree-species such as IBISCA - Panama 43

Brosimum utile (KUNTH) OKEN ex J. PRESL var. utile have attained a large size. The canopy crane stands in a six-hectare plot where all 21911 trees with a bole size of 10 mm in diameter at breast height (dbh) or greater have been identified (238 species), measured and mapped. Structurally, the forest can be characterized as having an average of 3676 stems per ha and a total basal area of almost 32 m2 per ha with the tallest trees reaching 45 m in height. Liana abundance amounts to 2222 individuals and 0.776 m2 basal area per ha (≥ 0.5 cm diameter; SCHNITZER, 2005). At least 103 epiphyte species are present within the crane perimeter (ZOTZ, 2004) and 119 liana and vine species have been recorded from the 6 ha plot (S.J. Fig. 1 — Monthly rainfall (bars) and average air WRIGHT, unpublished data). The most common plant temperature (line) for the study period species include Tovomita longifolia (RICH.) HOCHR., (September 2003 to November 2004) at San Protium panamense (ROSE) I.M. JOHNST., Tachigali Lorenzo (courtesy S. Paton, STRI). versicolor STANDL. & L.O. WILLIAMS and Psychotria suerrensis DONN. SM. in the understorey; and Brosimum utile var. utile, Aspidosperma spruceanum BENTH. ex MÜLL. ARG., Manilkara bidentata (A. DC.) A. CHEV. 2.2. Study sites and Tapirira guianensis AUBL. in the canopy. Overview The arthropod fauna of Panama is relatively well known, owing to many surveys and studies performed The IBISCA-Panama sampling protocols focused on 12 at Barro Colorado Island (BCI; QUINTERO & AIELLO, study sites, each 20 x 20 m in area. This 400 m2 ground 1992). Since BCI is ca. 25 km distant of the San Lorenzo surface area was chosen to match the projected area forest, we can expect some similarities in the fauna of of the canopy raft that would be sitting on the forest these two locations, although San Lorenzo is wetter than canopy (see Section 3). The 12 sites were situated in the BCI. Besides taxonomic and ecological studies listed surroundings of the STRI canopy crane (alt. 130m), on in QUINTERO and AIELLO (1992) and earlier references the top of a rather convex hill that slopes down towards collated therein, which may be relevant to the arthropod the banks of the Rio Chagres, with the exception of fauna of San Lorenzo, a number of ecological studies one site (R1), which was located in the floodplain. have also been performed within the San Lorenzo Sites were selected so as to best represent the variety forest. This trend has accelerated since the installation of the forest environment, with the distance between of the STRI canopy crane in 1997. In particular, recent the most remote sites being less than 2 km (Table studies focused on leaf miners and galling insects 1 and Fig. 2). To calculate distances between sites, (BARRIOS & MEDIANERO, 1999; MEDIANERO et al., geographic coordinates were converted to Cartesian 2003); chrysomelids and other phytophagous beetles coordinates by taking the relationship that 1° latitude (ØDEGAARD, 2003 [relevant checklist], 2006; CHARLES = 2 x pi x r / 360 and 1° longitude = cos(lat) = 2 x pi x & BASSET, 2005; ØDEGAARD et al., 2005; ØDEGAARD & r / 360, with r = 6372.8 km (mean earth radius) and lat FRAME, 2007); bees (ZAYED et al., 2003; WCISLO et al., = 9°16.771 (average latitude of sites). Distances were 2004); leaf-cutting ants (VILLELSEN et al., 2002; GERALDO then calculated using Pythagoras´ theorem (Table 1). et al., 2004: KWESKIN, 2004) and insect assemblages Sites were coded according to how the canopy was feeding on particular plants (BASSET, 2001; SCHOWALTER locally accessed (see Section 3, canopy access). Three & GANIO, 2003; ØDEGAARD, 2004). To this growing list, “crane sites” (C1, C2 & C3) were located inside the we can add some of the references listed in Section crane perimeter. Two “bubble sites” (B1 & B2) were 4 of the present article and numerous manuscripts in situated on either sides of the access road, ca. 1.5 km preparation, all resulting from IBISCA-Panama from the canopy crane. One “Ikos site” (I1), hosting the Ikos tree-house, was situated on a ridge not far from sites R2 and R3. Three “raft sites” (R1, R2 & R3) were far apart, with R1 located in the plain that borders Rio 44 Yves BASSET et al.

Plate 1 — [A] A selection of six plant species often encountered in the San Lorenzo forest, photographed from the canopy crane. (1) Guatteria dumetorum (Annonaceae), flowers and fruits; (2) branches of Poulsenia armata (MIQ.) STANDL. (Moraceae); (3) Marila laxiflora (Clusiaceae), flower; (4) tree crown of a large Manilkara bidentata (Sapotaceae); (5) Symphonia globulifera L. F. (Clusiaceae), flowers; (6) Pera arborea (Euphorbiaceae), branch with fruits. All photos by PC.

Plate 1 — [B] Insects from the San Lorenzo forest, representatives of various orders and feeding guilds. (1) Arachnoscelis sp. [Orthoptera Tettigoniidae, predator]; (2) Biolleyana costalis (FOWLER) [Hemiptera Nogodinidae, sap-sucker]; (3) Ptilotopus zonata (MOCSÁRY) [Hymenoptera Apidae, pollinator], buzz-pollinating flowers of the canopy tree Apeiba membranacea; (4) Cylindrotermes macrognathus SNYDER [Isoptera Termitidae, scavenger]; (5) Eciton burchelli WESTWOOD [Hymenoptera Formicidae, predator]; (6) Gibbifer impressopunctata (CROTCH) [Coleoptera Erotylidae, fungal-feeder]. Photos: ML (1-5), JS (6). IBISCA - Panama 45

Table 1 — Geographic coordinates of the 12 IBISCA sites; lower matrix of geographic distances; Simpson index of diversity calculated for each site; number of species expected in a sample of 100 individuals, S(100); and lower matrix of floral similarity as calculated with NNESS(100) (see text for details).

Site Latitude Longitude Geographic distance between sites (m) code (deg. N) (deg. W)

B1 9°17.133 79°59.106 B1 B2 C1 C2 C3 F1 F2 F3 I1 R1 R2 B2 9°17.146 79°59.290 B2 338 C1 9°16.774 79°58.495 C1 1303 1613 C2 9°16.793 79°58.499 C2 1279 1592 36 C3 9°16.779 79°58.468 C3 1342 1654 50 63 F1 9°16.453 79°58.642 F1 1521 1750 653 683 683 F2 9°16.472 79°58.590 F2 1548 1792 586 618 612 102 F3 9°16.482 79°58.582 F3 1543 1789 564 596 589 122 24 I1 9°16.617 79°58.377 I1 1644 1941 363 396 344 573 474 452 R1 9°17.322 79°58.533 R1 1108 1426 1018 983 1014 1623 1579 1560 1338 R2 9°16.769 79°58.320 R2 1592 1912 321 331 272 832 741 717 301 1097 R3 9°16.515 79°58.653 R3 1415 1653 561 588 596 117 140 144 540 1512 771

NNESS(100) between sites

Simpson S(100) B1 0.855 35.67 B1 B2 C1 C2 C3 F1 F2 F3 I1 R1 R2 B2 0.923 30.64 B2 0.333 C1 0.985 59.03 C1 0.416 0.397 C2 0.961 42.47 C2 0.400 0.494 0.509 C3 0.973 44.26 C3 0.329 0.528 0.624 0.586 F1 0.951 37.4 F1 0.407 0.541 0.551 0.561 0.601 F2 0.930 37.91 F2 0.419 0.500 0.486 0.622 0.602 0.699 F3 0.968 41.96 F3 0.375 0.494 0.565 0.624 0.674 0.663 0.680 I1 0.977 52.33 I1 0.439 0.494 0.554 0.619 0.604 0.543 0.638 0.599 R1 0.934 34.92 R1 0.187 0.114 0.207 0.175 0.182 0.124 0.150 0.146 0.199 R2 0.961 37.47 R2 0.395 0.515 0.426 0.570 0.575 0.534 0.566 0.620 0.593 0.061 R3 0.930 35.8 R3 0.309 0.467 0.555 0.531 0.587 0.621 0.619 0.654 0.573 0.164 0.525

Chagres, at a lower elevation than the other sites, whilst Botanical description of sites (PC, FH, AH, OP & R2 and R3 were located in the same area as sites I1 (and MS) also the fogging sites F1-F3). Site R1 was surveyed with the canopy raft, whereas sites R2 and R3 were At all sites, plants >10 mm dbh were tagged and equipped with climbing ropes and the canopy accessed identified, and their dbh measured (including Geonoma by single rope techniques. Finally, three “fogging sites” palms, that remain small at adult age) before any (see arthropod sampling, Section 4.3; F1, F2 & F3) were arthropod collection. Appendix I details species considered as “surrogate” sites for C1, C2 and C3, since abundance per site. Basal area was estimated by fogging was not allowed within the crane perimeter. computing pi x (dbh)2 / 4 for each tree and summing In the following sections, the authors responsible up the results for each site. A succession index was for the botanical and arthropod protocols detailed in the computed as the average of the specific fraction of text are identified by their initials in parentheses, and recruits (sfr) in light gaps x number of conspecific listed in alphabetical order. individuals in each plot, where the sfr parameter was obtained from WELDEN et al. (1991), for available species (60 out of 165). Table 2 summarizes the main characteristics of the sites. Average canopy openness at the sites was 7% (Table 2 and see Fig. 7.15). 46 Yves BASSET et al.

Fig. 2 — Map of the lower Rio Chagres, with the twelve IBISCA sites, coded as in Table 1. Upper inset: view of the Panama isthmus; lower inset: San Lorenzo protected area. The location of the study sites is indicated with a star.

“Crane sites” C1, C2, and C3. The largest trees were “Fogging sites” F1, F2, and F3. As in the C1-C3 sites, specimens of Brosimum utile var. utile. Other large trees the largest trees were Brosimum utile var. utile. Other of the three sites included specimens of Andira inermis large tree species included Aspidosperma spruceanum, (W. WRIGHT) DC., Cespedesia spathulata PLANCH., Inga Carapa guianensis AUBL., Guatteria dumetorum R.E. pezizifera BENTH., Jacaranda copaia (AUBL.) D. DON, FR., Eugenia coloradoensis STANDL., Humiriastrum Marila laxiflora RUSBY, Sloanea aff. meianthera DONN. diguense STANDL., Marila laxiflora, Tovomita longifolia, SM., Tabernaemontana arborea ROSE, and Vochysia and Virola multiflora (STANDL.) A.C. SM. The most ferruginea MART. The most abundant tree species were abundant tree species were the palm Socratea exorrhiza Marila laxiflora, Perebea xanthochyma H. KARST., (MART.) H. WENDL., as well as Dendropanax arboreus Tovomita longifolia, T. stylosa HEMSL. and Unonopsis (L.) DECNE. & PLANCH., Tovomita longifolia, Unonopsis panamensis R.E. FR. The understorey was dominated by panamensis and Xylopia macrantha TRIANA & PLANCH. the palms Geonoma congesta H. WENDL. ex SPRUCE and Geonoma congesta and G. cuneata were widespread in G. cuneata H. WENDL. ex SPRUCE. the understorey. IBISCA - Panama 47

Table 2 — Characteristics of the 12 IBISCA-Panama sites: number (No.) of stems (including those of a few unidentified trees); number of identified tree species; basal area; succession index (see text); canopy openness (proportion of visible sky (open) to closed in hemispherical pictures, analyzed by Hemiview 2.0); mean reflected light (average from 28 measurements within each site, cell facing down at dbh, measured with a digital lux meter LX-1010B, Kaito Electronics Inc., Montclair, USA); largest trees and most common species (*); dominant species with basal area (BA) or percentage occurrence (*); herbivory (mean percentage of leaves in samples with over 10 % of leaf area damaged, ± s.e.).

Site No. stems No. tree Basal area Succession Canopy Mean reflected light Largest trees / Dominant sp. Herbivory code >10mm species (m²) index openness (%) (lux, n=28) s.e.m. Most common spp. BA (m²), %

B1 211 51 1.01 14.13 7.55 7.96 2.19 Sap, Cal / Psy, Ing Cal 0.46, 45 36.2 ± 5.6 B2 127 34 1.62 17.39 6.00 57.18 10.14 Ape, Per / Bro, Geo Ape 0.34, 21 n/a C1 141 71 1.21 19.62 6.80 57.57 25.36 Bro, And / Prb, Geo Bro 0.68, 72 29.8 ± 4.2 C2 104 43 1.05 14.72 7.60 68.21 10.66 voc, Bro / Prb, Mar Voc 0.15, 23 38.1 ± 5.0 C3 121 49 1.16 15.91 10.60 8.46 1.60 Bro, Jac / Tov, Geo Bro 0.47, 46 39.4 ± 5.9 F1 116 40 1.41 13.83 7.10 n/a n/a Car, Mar / Geo, geu Car 0.29, 41 n/a F2 140 45 1.75 13.88 7.70 n/a n/a Bro, Vir / Geo, Geu Bro 0.53, 44 n/a F3 151 50 1.51 16.32 7.60 n/a n/a Bro, Hum / Geo, Den Bro 0.58, 38 n/a I1 124 59 1.13 16.33 6.90 21.61 5.05 Ape, Cal / Soc, Tov Ape 0.23, 20 41.8 ± 4.3 R1 123 39 2.93 17.64 2.91 26.56 1.92 Lue, Spon / Rin, Far Lue 1.08, 37 n/a R2 134 43 0.98 15.67 6.70 14.96 1.97 Asp, Man / Pro, Tov Asp 0.46, 47 n/a R3 107 37 1.03 16.60 6.90 44.96 5.54 Car, Tap / Geo, Mar Car 0.40, 39 n/a

* Plant names abbreviations: And = Andira inermis; Ape = Apeiba membranacea; Asp = Aspidosperma spruceanum; Bro = Brosimum utile var utile; Cal = Calophyllum longifolium; Car = Carapa guianensis; Den = Dendropanax arboreus; Far = Faramea occidentalis; Geo = Geonoma congesta; Jac = Jacaranda copaia; Lue = Luehea seemannii; Man = Manilkara bidentata; Mar = Marila laxiflora; Per = Pera arborea; Prb = Perebea xanthochyma; Pro = Protium panamense; Psy = Psychotria horizontalis; Rin = Rinorea squamata; Sap = Sapium sp. ‘broadleaf’; Soc = Socratea exorrhiza; Spo = Spondias mombin; Tap = Tapirira guianensis; Tov = Tovomita stylosa; Vir = Virola multiflora; Voc = Vochysia ferruginea.

“Bubble sites” B1 and B2. Site B1 showed remains Calophyllum longifolium, Dendropanax arboreus, of past constructions at ground level, that were not Simarouba amara AUBL., Tovomita longifolia, noticed when the site was chosen and that were and Virola sebifera. The best represented species evidently part of past jungle warfare training by U. S. included Brosimum utile var. utile, Matayba apetala soldiers (WEAVER & BAUER, 2004). The largest trees RADLK., Protium panamense, Tovomita longifolia, in B1 belonged to the species Sapium sp. “broadleaf” Virola sebifera, and the palm Socratea exorrhiza. and Calophyllum longifolium WILLD. Other large trees Geonoma congesta and G. cuneata were present in the included Terminalia amazonia (J.F. GMEL.) EXELL and understorey. Lacmellea panamensis (WOODSON) MARKGR. The most “Raft sites” R1, R2, and R3. Two trees at the R1 site abundant woody species was Psychotria horizontalis exceeded 1 m in diameter, one Luehea seemannii TRIANA SW., with Inga sertulifera DC., Piper colonense C. & PLANCH. and one Spondias mombin L. Anacardium DC. and Sorocea affinis HEMSL. being well represented. excelsum (BERT. & BALB. ex KUNTH) SKEELS and Carapa Only one specimen of Geonoma congesta was recorded guianensis also grew there as large trees. The most on the site, and no G. cuneata. The largest trees at the abundant species were Carapa guianensis, Desmopsis B2 site were an Apeiba membranacea SPRUCE ex BENTH. panamensis (B.L. ROB.) SAFF., Faramea occidentalis and a Pera arborea MUTIS. Tapirira guianensis, Laetia (L.) A. RICH., Oxandra panamensis R.E. FR., and procera (POEPP.) EICHLER, Brosimum utile var. utile and Rinorea squamata S.F. BLAKE. No Geonoma palms were Virola sebifera AUBL. also grew as large trees at this site. recorded from the understorey at this site. The largest The most abundant species were Brosimum utile var. trees at site R2 were several Aspidosperma spruceanum. utile, Perebea xanthochyma, Protium panamense and Other large trees included Brosimum utile var. utile, the palms Socratea exorrhiza and Oenocarpus mapora Guatteria dumetorum, Manilkara bidentata, Marila H. KARST. Geonoma congesta was well represented in laxiflora, Matayba apetala and Tapirira guianensis. the understorey, but G. cuneata was not recorded at this The most abundant species were Maranthes panamensis site. (STANDL.) PRANCE & F. WHITE, Marila laxiflora, Perebea “Ikos site” I1. The largest trees at this site belonged xanthochyma, Protium panamense, Tovomita longifolia, to Apeiba membranacea, Brosimum utile var. utile, and T. stylosa. Only one Geonoma congesta occurred at 48 Yves BASSET et al. this site, and no G. cuneata. The largest trees at the R3 site were Carapa guianensis and Tapirira guianensis. Brosimum utile var. utile, Cecropia insignis LIEBM. and Marila laxiflora also grew as large trees at this site. The most abundant species were Brosimum utile var. utile, Marila laxiflora, Perebea xanthochyma, Protium panamense, Xylopia macrantha and the palms Socratea exorrhiza and Synechanthus warscewiczianus H. WENDL. In the understorey, Geonoma congesta was more abundant than G. cuneata.

Vegetation analysis (YB, PC, OJH, ML)

The 12 IBISCA sites included 1,556 individual trees of more than 10 mm dbh (including also adult Geonoma palms), representing 163 species, in a total area of 0.48 ha. The most abundant species was Geonoma congesta (129 individuals), followed by Perebea xanthochyma, Psychotria horizontalis and Brosimum utile var. utile (69, 69 and 65 individuals, respectively). The most diverse family was Fabaceae s.l. with 16 species, and three other families included ten or more species (Annonaceae: 10 spp., Arecaceae: 12 spp., Rubiaceae: 11 spp.). In terms of number of individuals, Arecaceae was the most abundant family (315 individuals, of which 175 belonged to one of the two Geonoma species). Four other families were represented by more than one hundred individuals (Annonaceae: 119 ind., Fig. 3 — Comparison of tree species richness (individual- Clusiaceae: 179 ind., Moraceae: 174 ind., Rubiaceae: based rarefaction, COLEMAN method) and 121 ind.). Locally abundant species are detailed in the abundance in the twelve 400m² plots investigated description of sites and in Table 2. during the IBISCA-Panama project. A few The study sites were chosen in order to examine unidentified trees were not taken into account in this analysis. arthropod beta-diversity, and hence were not specifically designed to investigate vegetation changes according to distance. However, the floristic composition of sites Second, data on plant species-abundance at each site is expected to be of prime importance in determining were compiled to build pair-wise measures of similarity the distribution patterns of arthropod abundance and between sites. For this purpose, we calculated two diversity. Hence, the following four analyses aimed quantitative indices of species similarity, the MORISITA- to compare plant distributions between sites. First, HORN index and the NNESS index, with software the software EstimateS (COLWELL, 2004) was used to provided by HARDY (2007). NESS is a metric relatively compute rarefaction curves of plant species richness insensitive to sample size but sensitive to rare and at each site based on individuals (COLEMAN method). abundant species; NNESS is a modification of NESS We also computed Simpson diversity index for each to allow calculations with singletons. NNESS values site and the expected number of species occurring in a were calculated with sample size parameter k set to sub-sample of size k, with k representing the maximal 1 (identical to MORISITA-HORN index, most sensitive value permitted to compare all pairs of samples (k = 100 to common species), to the maximal k for comparing in our case), with software provided by HARDY (2007). all pairs of samples (most sensitive to rare species; Species rarefaction indicated that site C1 was the richest in our case k = 100, hence the notations S(100) and and most diverse site, followed by I1, C3, C2, F3, F2, NNESS(100) in Table 1; GRASSLE & SMITH, 1976; HARDY, F1, R2, B1, R3, R1 and B2 (Fig. 3). Site B1 was the 2007). Both the plots of MORISITA-HORN similarites most densely populated site with trees (Fig. 3, Table 1). (not presented here, R2 = 0.322, MANTEL test p=0.026) IBISCA - Panama 49

Fig. 4 — Scatter plot of pair-wise NNESS(100) tree species similarity between sites (vertical axis) plotted against straight- line geographic distances between sites (horizontal axis; metres), detailed for different pairs of sites (open squares = pairs including site R1; crosses = pairs including site R1; closed diamonds = pairs including other sites). The negative regression is significant both for pairs including all sites and for pairs with the exception of B1 and R1, as shown here (see text). and NNESS(100) similarities were significantly and tolerant shrub species does not recruit highly from light negatively correlated with geographic distances (Fig. gaps (WELDEN et al., 1991), but it can be propagated 4, R2 = 0.334, MANTEL test p=0.018, including all pairs easily in the forest from sterns, leaves, or leaf fragments of sites). The similarities of B1 and R1 with other sites (SAGERS & COLEY, 1995). Hence, the past constructions were low, but the correlation between similarity and at ground level observed at site B1 seem overall to distance remained similar when pairs including B1 and have little altered the floral composition of site B1 and R1 are removed from the analysis (Fig. 4, R2 = 0.335). its canopy openness (as evidenced by low succession Third, the matrices of geographic distance and index and canopy openness values of site B1 in Table floral similarity (MORISITA-HORN index) were used to 2), but we cannot discount that the local spread of compute a dendrogram using WARD’s algorithm (Fig. P. horizontalis has been promoted by this past 5). This analysis confirmed the previous one and, in disturbance. On the other hand, site R1 was clearly particular, the low floristic similarities of sites B1 and floristically different from other sites, with several plant R1 with other sites. Site B1 was, overall, not much species not shared with other sites, due the location different floristically from other sites but was heavily of this site in the floodplain of the Rio Chagres. The dominated in frequency by a single species, Psychotria dendrogram also suggested that sites F1-F3 could not be horizontalis, absent from other sites. This shade- considered as true surrogates of sites C1-C3 (Fig. 5). 50 Yves BASSET et al.

was averaged for species occupying the same rank in different samples). HUBBELL (2001, p. 150) used a similar approach, with the pooled sites considered a surrogate for the meta-community, and the ‘averaged sites’ used as an estimate of dominance-diversity curves at a smaller scale. Fig. 6 shows the curves obtained for the IBISCA sites and the 6 ha plot. The curve for the pooled sites was not too different from the 6 ha plot curve, as was expected from two samples closer to the meta-community than individual plots. On the other hand, the averaged curve looked distinctly steeper (i.e. the rarest species in the plots appeared on average rarer than in the approximated meta-community). A similar pattern was found by HUBBELL (2001), who attributed it to dispersal limitation at a small scale, a reason that is possibly applicable to the IBISCA sites as well.

3. Canopy access (Plate 2)

During IBISCA-Panama, different techniques (canopy fogging, single rope techniques) and devices (canopy crane, canopy raft, canopy bubble and Ikos) allowed field participants to survey arthropods in the upper canopy, either in situ or indirectly. This was the first time that these complementary techniques and devices had been deployed in parallel (ROSLIN, 2003). Canopy fogging may be considered as a sampling technique, yielding indirect access to the canopy Fig. 5 — Cluster analysis (WARD’s method) of similarity (see Section 4.3). Single rope techniques (SRT) are in geographic distances between sites (A) and undoubtedly the cheapest way to reach the canopy. with MORISITA-HORN similarity of plant species Although SRT requires training and a reasonable level composition between sites (B). of physical fitness, it is used by increasing numbers of canopy researchers (BARKER & STANRIDGE, 2002). A Fourth, HUBBELL (2001) discussed the possible minority of IBISCA-Panama participants were trained signification of dominance-diversity curves in the climbers. Most of participants relied on the assistance framework of his neutral theory of biodiversity and of professional climbers, but used SRT to access the biogeography. From various simulations, it appears canopy raft and Ikos. Two sites were mainly accessed that steep curves tend to be associated with early with SRT. states that are still far from dynamic equilibrium, Canopy cranes are currently the most appropriate or with systems having low θ values (a biodiversity devices allowing long-term research on canopy parameter encapsulating the size of the source biodiversity (BASSET et al., 2003b). The Smithsonian community - a.k.a the meta-community - and the rate Tropical Research Institute first developed the idea of of speciation), whereas S-shaped curves are indicative using construction cranes to access the forest canopy of systems close to equilibrium and with high θ values. in Panama (PARKER et al., 1992). The first permanent Dominance-diversity curves are obtained by ranking canopy crane was installed in 1992 in the Parque species in order from the most abundant to the rarest, Natural Metropolitano, near the capital of Panama. In and plotting the log of each species relative abundance 1997, STRI installed a second crane in the San Lorenzo against its rank. The curves were computed for (a) Protected Area (WRIGHT et al., 2003). A metal basket the 6 ha plot (see description earlier in the text), (b) called the gondola, in which observers stand, is hoisted all IBISCA sites pooled together and (c) the IBISCA above the forest by the crane and lowered to research sites averaged (the relative abundance of species locations within the canopy. Canopy cranes enable IBISCA - Panama 51

Fig. 6 — Dominance-diversity curves for the plant species of the 12 IBISCA-Panama sites, averaged and pooled, compared with the curve obtained for the 6 ha plot next to the San Lorenzo canopy crane.

easy and safe three-dimensional access to the canopy. tree tops within reach. Compared with canopy cranes, The movement of the crane and gondola is controlled the mobile canopy raft allows comparisons between through radio communication with a crane operator. remote sites. Usually an air-inflated dirigible raises The tower height of the San Lorenzo crane is 52 m, the raft and sets it upon the canopy. During IBISCA- its jib length 54 m, the maximum height reached by Panama a Bell-212 helicopter from the Servicio Aero the gondola is 49.5 m and the area of accessed forest Nacional was used for this purpose. The helicopter 0.92 ha. During IBISCA-Panama this crane was used to carried the canopy raft at the end of an array of access the upper canopy of three sites. ca. 60 m cables, in order to cause minimal disturbance The SolVin-Bretzel canopy raft corresponds to the to the canopy. Unfortunately, due to unexpected 4th generation of canopy raft (“radeau des cimes” in maintenance problems with the helicopter, the canopy French; HALLÉ, 2000). It is a 400 m2 platform of 500 kg, raft was only installed at one site. shaped like a pretzel and consisting of air-inflated beams The canopy bubble (or “bulle des cimes” in French) is and Aramide™ (PVC) netting that rests on the top of an individual 180 m3 helium balloon of 6 m in diameter canopy trees. It was first used in the Masoala National which runs along a fixed line set up in the upper canopy. Park (Madagascar) in 2001 and for the second time The observer is seated in a harness suspended below the during IBISCA-Panama. Access to the raft is provided balloon and moves along the line by hands or with the by single rope techniques. Researchers can walk on aid of mechanical jumars. The canopy bubble was first the net from one side to the other and may collect used in Gabon (1999) and Madagascar (2001). One such specimens by leaning over and collecting from the device is permanently installed in the Nouragues station 52 Yves BASSET et al.

Plate 2 — Canopy access devices, clockwise: SolVin-Bretzel canopy raft, canopy bubble, Ikos tree house and canopy crane. Photos: S. Bechet, R. Le Guen, JO, ML. IBISCA - Panama 53 in French Guiana, with a 5 km guide-rope network. In below the scope of each arthropod sampling protocol Panama the guide ropes were installed by helicopter and (Fig. 7) and Table 3 summarizes the sampling effort for allowed access to the upper canopy of two sites. each protocol. In total, 1386 traps and other protocols The Ikos is an icosahedral unit (i.e., twenty sides), produced 9402 samples, equivalent to 24354 trap-days 3.2 m in diameter, installed in the fork of a large tree, (or person-days) of sampling effort. A substantial part of which is designed primarily for long-term observations this sampling effort focused on the upper canopy of the (a canopy tree-house). It can be assembled in the tree forest. The spatial replication achieved with protocols or partially built on the ground and then hoisted up. such as light, flight-interception and sticky traps, Single rope techniques provided access to the Ikos. and BERLESE-TULLGREN was high and has little or no Researchers can work on a platform, which is attached equivalent in the published literature, particularly when to the top of the structure. The Ikos was first used in the considering the vertical dimension. The advantages and Forêt des Abeilles (Gabon) in 1999 (HALLÉ, 2000) and limitations of each sampling protocol are discussed in then in Masoala National Park (Madagascar) in 2001. SOUTHWOOD (1978), MUIRHEAD-THOMSON (1991) and During IBISCA-Panama the Ikos allowed access to the BASSET et al. (1997). upper canopy of one site. 4.2. Soil and litter samples WINKLER sifting (HPA, BC, AD, ML, JO & YR) 4. Outline of arthropod sampling protocols 4.1. Timeframe, seasonal replications and overview This method collected active and passive arthropods in ground litter. The leaf litter was sifted and then extracted A first six-week field study during the late wet season for 48 hours using a mini-WINKLER apparatus (as detailed of 2003 (22 September to 31 October) concentrated in FISHER, 1998) to obtain all arthropods. In September- on all 12 sites, using 14 arthropod protocols. It October 2003, 8 sites were surveyed, from which 51 was attended by 45 participants from 15 countries, samples of 1 m2 of leaf litter were extracted per site. In including entomologists, botanists, students (University May 2004, three sites were re-surveyed, yielding 153 of Panama, Maestria de Entomologia) and 10 technical additional samples of 1 m² of leaf litter. Such sampling staff (professional tree climbers, consultants and intensities yield a representative estimate of the local photographers). Repeated seasonal replicated sampling structure of ant assemblages (LEPONCE et al., 2004) in 2004 included shorter periods of field work, focusing on a restricted number of study sites with a restricted BERLESE-TULLGREN (GCM, NNW) number of arthropod protocols: This method collected microarthropods in ground soil Seasonal replicate 1: 1 February- 15 March 2004: five and in suspended debris accumulations in the canopy. weeks; dry season; 7 participants who concentrated on At each site, three trees were sampled and 16 cores 2-6 sites with five protocols. (3 x 5 cm each) were obtained from each tree: 8 from the ground and 8 from suspended debris accumulations Seasonal replicate 2: 10-31 May 2004: three weeks; in the tree. Each soil core (sample) was extracted for onset of the wet season; 27 participants (including two 48 hours in a BERLESE-TULLGREN apparatus and the parataxonomists) who concentrated on 2-8 sites with 14 resulting arthropod specimens were collected into 75 % protocols. ethanol (see WINCHESTER & BEHAN-PELLETIER, 2003 for more details). In 2003, 8 sites were surveyed, yielding Seasonal replicate 3: 15 October - 22 November 384 samples. In May 2004, 4 sites were surveyed, 2004: five weeks; late wet season; 9 participants who yielding 176 additional samples. Sub-samples of the concentrated on 2-6 sites with six protocols. Collembola material were studied by CASTAÑO-MENESES et al. (2006). Thus, seasonal replicates were conducted three times at least at the three crane sites. One full replication Hand collecting: social insects (BC, AD, ML, JO & (all sampling protocols) took place in May 2004. YR) Arthropods were collected from inside the 400 m2 area of each site after completion of the vegetation survey, Hand collecting for social insects (ants and termites) unless sampling protocols necessitated a larger surface was performed on the ground, in the understorey and in (e.g., hand collection of ants and termites). We detail the mid- and upper canopy, with substantial help from

56 Yves BASSET et al.

professional climbers. In 2003, 8 sites were surveyed, for a total sampling effort of 630 trap-days, obtained yielding 160 samples of 10 m2 on the ground and 45 from seven sites (MEDIANERO et al., 2007). samples in the canopy (termites); as well as two 170 x 40 m transects and approximately 300 samples 4.3. Understorey and/or canopy samples in the canopy and understorey each (ants). In 2004, Canopy fogging (JB, AF, & JS) ground termites were collected from 40 quadrats of 5 m² located at intervals of 10 m along a transect line A wide range of arthropods was collected with this (yielding approximately 100 samples). Canopy samples method, which consisted of dispersing knockdown were collected from 40 emergent trees along a transect insecticide up into trees with a fogging machine line (n = 32 samples collected); arboreal termites were operated from the ground or from a tree. The stunned collected from nests, galleries and pieces of dead wood arthropods fell on to collecting trays installed at 1 m observed in the canopy; and arboreal ants were collected height above ground (for more details, see ADIS et from live branches (approximately 100 samples; ROISIN al., 1998 and KITCHING et al., 2002). A Swingfog ® et al., 2006). SN1 machine was used with a 1 % natural Pyrethrum solution, dissolved in white oil Essobayol ® 82. At each Pitfall traps (EM & AT) fogged site, six samples were obtained, corresponding to the arthropods that fell onto six distinct 4 x 5 m Pitfall traps collected active arthropods in the ground plastic sheets installed at 1 m height. In 2003, eight litter. Traps were made of plastic cylinders of 6 cm in sites were fogged (48 samples), whereas six sites were diameter and 15 cm depth (0.424 l) containing a solution re-fogged in May 2004 (36 samples) and October 2004 of 50 ml ethanol, 10 ml of liquid detergent and 5 g salt. (36 samples). Traps were covered with an 18 cm raised plastic plate to protect them from rainfall. At each study site, 15 traps Beating of vegetation and dead branches (HB & FØ) were buried flush with the soil surface, in a straight line at 1.3 m intervals. Traps were surveyed after three days, Beating collects passive foliage arthropods and was

Table 3 — Sampling effort for each arthropod protocol: number (no.) of traps (if relevant), no. of sites surveyed, no. of seasonal replicates, types of habitats surveyed (s = soil, l = litter, u = understorey, m = mid-canopy, c = upper canopy), no. of samples, no. of trap-days (or person-days, whichever is relevant), and no. of arthropod individuals collected. NA= not available. Protocol# Traps Sites Replicates Habitats Samples Trap-days Individuals (x1000)

WINKLER — 8 2 l 561 44 24 BERLESE TULLGREN — 8 2 sc 528 22 30 HC: social insects — 10 2 lumc 848 129 1.5 Pitfall traps 210 7 4 l 210 630 10 Fogging — 8 3 m 120 19 81 Beating — 8 4 uc 560 11 3.5 Baits — 8 2 uc 59 28 1.3 HC: galls — 5 2 umc 231 14 5 Wood rearing — — 2 uc 2303 1213 NA Light traps 48 8 4 uc 90 90 55 Aerial FITs 90 5 4* lumc 1659 16244 69 Ground FITs 36 9 2 u 173 248 15 Sticky traps 993 8 4 lumc 1986 4965 56 Malaise traps 9 9 3 u 74 697 165

TOTAL 1386 12 4* slumc 9402 24354 >500

#HC = hand collecting; FIT = flight-interception trap *Aerial FITs ran continously from October 2003 to October 2004 IBISCA - Panama 57 performed during day-time either on live vegetation or recorded: height of branches; number of leaves; number dead branches. For the former, arthropods were collected of dead or active insect galls; number of leaves with on square beating sheets of 0.4 m2 in area, of conical more than 10 % of leaf area lost by chewing insects; and shape, ending in a circular aperture (7 cm in diameter), number of active meristems. Five sites were surveyed which was fitted with a removable plastic bag. Sheets in 2003 and re-surveyed in May 2004, totaling 43994 were inserted below the foliage so that one layer of leaves inspected in 231 branch samples from 73 tree leaves above occupied approximately the entire area of and liana species (RIBEIRO & BASSET, 2007). the sheet. Arthropods were dislodged from the foliage with three good strokes, and gently brushed inside the Wood rearing (HB, LC, GC, FØ & MR) plastic bag, which was then closed and replaced by a new one. A total of 560 beating samples were obtained Xylophagous insects were reared from baits of cut from the vegetation at 7 sites, half from the understorey branches of 15 tree species, collected from outside the and half from the upper canopy (sites C1 and C2 were study sites. All timber baits included about 15 kg of replicated in February, May and October 2004). fresh wood. For each tree species, a minimum of eight To compare the vertical stratification of saproxylic timber baits were placed as follows, shortly after being beetles within and between different plant species, cut: two baits in the shade, understorey; two baits in the another protocol was used. Cut branches of 16 tree sun, understorey; two baits in the shade, upper canopy; species were first suspended in their parent tree. and two baits in the sun, upper canopy. After one month For each tree species, four branches were placed in of exposure to ovipositing females of saproxylic insects, the canopy (15-25 m above ground) and four in the all baits were removed and enclosed in emergence understorey (1 m above ground). All branches were bags of fine black mesh, fitted with a ‘Malaise-type’ beaten regularly, approximately every second day over collecting bottle at the upper corner, and filled with a three-week period starting five days after branches a 5 % formaldehyde-ethylene glycol solution. Pilot were cut. A total of 304 and 268 samples from the studies were initiated in 2004 but all samples (n = 2303) understorey and canopy were obtained, respectively. All were collected during additional field work in 2005. To Coleoptera associated with dead wood and senescing date, not all reared specimens have been processed and leaves were collected. This protocol was initiated in databased. 2003 and replicated in May 2004. Light traps (AC, RLK, MMa, MMo & EGO) Baits and netting (DMF & DWR) Our light traps surveyed insects that are attracted Using honey water as bait for meliponines and Cineol™ to light, including moths. We used a commercially for euglossines, bees were netted in the understorey available design based on the so-called Pennsylvania (or (approximately at breast height) and in the upper canopy Texas) trap (FROST, 1957). Essentially this comprised a at eight sites in 2003 and six sites in May 2004. The baits vertically-mounted ‘black light’ fluorescent tube with were left for two hours at each location, meliponines three transparent plastic vanes mounted equidistantly were sampled and euglossines systematically caught. A around it. These vanes were shaped to fit within a total of 59 samples were obtained. funnel and the funnel sat within a replaceable bucket in which the catch accumulated. The light operated using a Hand-collecting: gall-forming insects and herbivory 12 Volt gel battery. To this commercially available model (SPR) we added a rain protector of a size larger than the bucket (we used an alloy dustbin lid around 60 cm diameter) This protocol aimed at quantifying galling performance and a sandwich of wooden boards beneath in which and free-feeding herbivory. It consisted of sampling the battery could be mounted. These modifications galls on leaves observed within the volumetric space ensured that: (a) the trap could be used in wet to very of a vertical cylinder 1 m diameter, established from wet conditions; and (b) the trap and its battery could the upper canopy to the understorey (up to three be hauled into the canopy by rope. We used a block meters above the ground). At each site, three vertical of Dichlorvos™ impregnated plastic as a killing agent “transects” were chosen randomly, within the range of placing it in the bucket together with torn baking paper reachable spaces. An additional horizontal 30m-long to provide resting places for captured insects (for more transect was surveyed in the understorey of each site. details see KITCHING et al., 2000). Within each transect, the following measurements were The upper canopy traps were hung from tree branches 58 Yves BASSET et al.

(ca. 25-35 m above ground). The understorey traps were sided), yellow-colored, plastic cardboard coated hung at approximately 2 m above ground from ropes with glue (Kollant Temo-O-Cid Colortrap, Kollant tied to two adjacent tree trunks. Each site was equipped Padova, Italy). At each site, 25 traps were placed in the with three canopy and three understorey traps, which understorey (at dbh, 1.3 m) and 25 traps in the upper were run for a night. The traps were usually set at canopy (treetops, usually 30-35 m). In addition, three about 5-6:30 PM (sunset) and removed the following vertical transects, consisting of seven traps placed at morning around 6-7:00 AM (sunrise). In 2003, eight 0 m, 1.3 m, 7 m, 14 m, 21 m, 28 m and the treetops sites were surveyed (48 samples). In 2004, sites C1 and (upper canopy), were surveyed at each site. Thus, in C2 were re-surveyed in February, May and October total 71 traps were set up at each site and each left for (36 samples). An additional site was also surveyed in five days. Overall, nine sites were surveyed and 993 May 2004 (6 samples). traps were used, with sites C1 and C2 being replicated three times in 2004. In addition, sticky traps were also Aerial composite flight-interception traps (RKD, LLF installed specifically to collect wood-boring Agrilus & MR) (Coleoptera, Buprestidae) near dead trees in the study area (CURLETTI, 2005; CURLETTI et al., 2006). These traps collected large flying insects in the understorey and canopy. Each consisted of a rain cover Ground Malaise traps (SP & Neil D. Springate) and two 60 cm tall, perpendicular perspex intercept sheets above a 23 cm diameter collecting funnel and Our Malaise traps collected various insects flying in the preserving jar (active collecting area = 0.55 m2, double understorey. We used a model from Harris House Nets sided). At each study site, three vertical trap transects (B&S Entomological Services, Portadown, N. Ireland) were installed. Each transect consisted of six traps set with an active collecting area of ca. 4 m2 (double-sided), up at different heights (0 m, 1.3 m, 7 m, 14 m, 21 m and based on the basic design of TOWNES (1972). Malaise 28 m). Ground traps (0 m) were placed near the base of traps were emplaced on the ground at nine sites in 2004 each tree and were dug into the ground level with the and later reduced to six sites, due to human interference. top of the collecting funnel, and in this way they acted Traps were surveyed every 10 days during three seasonal as combined pitfall and flight traps (for more details replicates in 2004, yielding 74 samples. see EWERS et al., 2007). Traps were usually surveyed every ten days. They ran more or less continuously from 4.4. Processing of material October 2003 to October 2004 on 4-5 sites, yielding 1659 samples. During IBISCA-Panama, each participating entomologist in the field was responsible for a particular Ground flight-interception traps (AT) sampling protocol and the study of 1-2 focal taxa (= ‘supervisor’). Initial sorting at the ordinal or familial These traps were erected on the ground as large level was facilitated by students from the University rectangular fences (3 x 1 m) made of fine plastic mesh of Panama and by parataxonomists from the New (active collecting area = 6 m2, double-sided). They Guinea Binatang Research Center (for a discussion collected large flying insects, which impacted the fence on parataxonomist duties in biodiversity projects, see and fell into the collecting trays disposed below the BASSET et al., 2004). The supervisors collaborated fence, which were filled with water and ethanol. Three with taxonomic authorities for the formal study of the such traps were run per site, each for an exposure period material (88 additional participants), first pre-sorting of 1-2 days. Eight sites were surveyed in 2003 and May the material into morphospecies, then identifying them 2004, yielding 120 samples. In addition, traps were run as far as current knowledge allowed. In total, IBISCA- at other locations in the study area as part of a study of Panama will evaluate patterns of spatial distribution for myrmecophilous Histeridae (Coleoptera) and yielded 63 focal taxa, representing different lineages and life an additional 53 samples. history strategies (Table 4). Thorough taxonomic study of this material is Sticky traps (HPA, HB, YB, LC & GC) essential for at least two reasons. First, morphospecies need to be cross-checked both among study sites and Sticky traps collected small flying insects in the between levels within the forest (from the soil to the understorey and canopy. They consisted of small (29 upper canopy) in order to be able to include them in x 12.5 cm, active collecting area = 0.0725 m2, double- cross-site comparisons and beta-diversity estimates. IBISCA - Panama 59

Second, the Latin binomials of identified species provide been named and described, and how widely are they access to additional ecological information available distributed! Perhaps one of the main reasons for this from the published literature. For specific focal taxa, is that sufficient primary research has been conducted the entomological fauna of Panama is reasonably well to a point where global cycles of gases and nutrients known, in comparison with other tropical countries, so can now be relatively well explained by extrapolation that information can be extracted from a relatively large from observations at smaller-spatial scales. In contrast, literature. the magnitude and determinants of biodiversity are far A special database with a MS-Access 97 core from being reasonably well understood. For example, was developed in order to register conveniently all knowledge of local food webs, especially in the tropics, arthropod records. The definition of a common set is still rudimentary (GODFRAY et al., 1999; KITCHING, of codes for samples, higher taxa (i.e., taxa ranking 2006). In short, the inability of the scientific community above genus level), guilds and host plants prevented to document species diversity, let alone alterations problems of duplication. The individual data files brought about by global environmental change, is from each supervisor responsible for a taxonomic hugely detrimental to the credibility of the conservation group or protocol were merged in a master database. movement (MANN, 1991). The magnitude of the effort This collective data matrix summarizes interactions and cost necessary for adequate surveys of biodiversity at eight main sites between five main habitats (soil, and implementation of conservation policies may also litter, understorey, mid-canopy and upper canopy), be overwhelming to many individuals, organizations 14 sampling protocols, 9402 samples, four seasonal and institutions. replicates, 315 plant species. We estimate that the final The new modus operandi of IBISCA involves product should shed light on the spatial and seasonal team work (ecologists, taxonomists, students, distribution of approximately half a million specimens parataxonomists), international collaboration and belonging to ca 5500 species, distributed among 63 complementary skills, both in the field and laboratory. focal groups of different phylogeny and ecology. No There have been major multi-scientist studies of tropical comparable dataset exists for arthropods of tropical nature before, but these have seldom been coordinated rainforests. within a strong, single experimental design that will Many species collected during IBISCA-Panama generate an accurate picture of the spatial distribution appear to be new to science and will be described in due of arthropods within the study area. Specifically, with course. So far, 26 new species have been described from IBISCA-Panama, it will be possible to compare, for tenebrionid and buprestid material alone, including one the first time, faunal samples obtained in situ from the species named in honour of the programme, Lenkous canopy with those obtained from the understorey or soil- ibisca FERRER & ØDEGAARD (CURLETTI, 2005; FERRER & litter of a tropical rainforest with sufficient taxonomic, ØDEGAARD, 2005). New species often included specimens spatial and temporal replication. Such data were originating from the upper canopy, but not always. New collected in the field at a fraction of the cost associated species were often cryptic, such as the many species of with recent research initiatives in other life sciences. Anobiidae (Coleoptera) collected by fogging. Inevitably, All fieldwork-related expenses amounted to less than this probably represents but a fraction of the task awaiting US$ 300,000. Although acquiring raw data in the field taxonomists. The material collected by IBISCA-Panama was relatively inexpensive, the costs of processing, may also be instrumental in documenting the spread of identifying and databasing IBISCA specimens may alien species, as shown by KIRKENDALL and ØDEGAARD easily double the overall expenses of IBISCA-Panama. (2007) for some scolytine species. Still, this investment appears low in comparison to budgets of large-scale genetic research programmes, and a small price to pay to stimulate innovative lines of 5. Significance of IBISCA-Panama and perspectives research in tropical ecology. Multi-protocol, multi-habitat and multi-taxa studies, In some sense, it is puzzling to observe that scientists have such as IBISCA, are essential if statements are to be managed to focus public attention on complex problems made about the overall arthropod diversity of forests. of gas exchange and nutrient cycling involving many The assumption, tacit since ERWIN & SCOTT’s (1980) processes that are invisible to the naked eye, whereas article, that the species richness of the forest is totally they have largely failed to focus public attention on the canopy-dominated is certainly not true. IBISCA- fundamental and unanswered questions of how many Panama may be considered as a model for large-scale species inhabit the Earth, what proportion of these have investigative programmes focusing on arthropod 60 Yves BASSET et al.

Table 4 — Current status of taxonomic knowledge for the major focal taxa studied by IBISCA-Panama, as of August 2007: order; feeding guild; supervisors (abbreviated by initials); number of individuals databased (Ind.), total number of species or morphospecies databased (Spp.), percentage of species so far identified to species level (Id.). To put these preliminary data in perspective, the estimated number of known species from Panama is also indicated (Panama), compiled from various sources. In total 63 focal taxa are studied under the responsibility of 27 supervisors and 109 taxonomists.

Focal area Order Guild Supervisor(s) Ind. Spp. Id. Panama %

Oribatei Acari varia NNW >16,000 >139 33 ca. 200 Araneae Araneae Predator JS >5,000 >200 0 - Opiliones Arachnida Predator JS >400 36 0 - Ricinulei Arachnida Predator JS 17 2 0 6 Collembola Collembola Scavenger GCM 11,603 50 4 23 Curculionoidea Coleoptera varia HB ca. 7,000 548 35 ca. 4,000 Cucujoidea/Tenebrionoidea (part) Coleoptera varia LC 2,000 210 0 - Chrysomelidae (part) Coleoptera Chewer AF >400 >115 21 - Cerambycidae Coleoptera Wood-eater FØ >150 57 68 - Ceratocanthidae Coleoptera Fungal-feeder AT 313 24 15 40 Histeridae Coleoptera Predator AT 1648 173 33 250 Pselaphidae Coleoptera Predator AT 1402 142 22 508 Scolytinae & Plarypodinae Coleoptera Wood-eater FØ >5,000 144 34 - Scarabaeoidea Coleoptera Scavenger FØ >300 48 96 541 Scydmaenidae Coleoptera Predator FØ >250 63 3 - Buprestidae: Agrilus Coleoptera Wood-eater GC 44 22 18 28 Nitidulidae Coleoptera Varia AT ca. 2,500 45 35 92 Erotylidae & Endomychidae Coleoptera Fungal-feeder LC & JS >500 103 0 - Cleridae Coleoptera Predator JS >600 46 0 - Eucnemidae Coleoptera Wood-eater JS >500 79 0 - Elateridae Coleoptera Various JS >1,000 99 28 - Empididae Diptera Predator RKD & LLF >300 - 0 - Scatopsidae Diptera Various RKD & LLF >100 - 0 - Mycetophilidae Diptera Fungal-feeder RKD & LLF >100 - 0 - Phoridae Diptera Various RKD & LLF >5,000 - 0 - Sphaeroceridae Diptera Various RKD & LLF >2,000 - 0 - Ceratopogonidae Diptera Various RKD & LLF >8,000 - 0 - Dolichopodidae Diptera Predator RKD & LLF >2,000 - 0 - Drosophilidae Diptera Various RKD & LLF >2,000 - 0 - Membracoidea Hemiptera Sap-sucker YB 1460 57 58 137 Fulgoroidea Hemiptera Sap-sucker YB 5400 180 31 500-1,000 Psylloidea Hemiptera Sap-sucker YB 1953 27 59 ca. 150 Cicadellidae Hemiptera Sap-sucker YB 3827 160 33 - Meliponini & Euglossini Hymenoptera Pollinator DMF & DWR 1363 42 / 22 98 ? / 35

Braconidae Hymenoptera Parasitoid EM 2,500 ca. 80 31 - Formicidae Hymenoptera Ant BC, AD, ML & JO ca. 50,000 291 57 301 Isoptera Isoptera Scavenger YR & ML ca. 15,000 60 55 50 Geometroidea Lepidoptera Chewer RLK >313 >108 77 ca. 1,500 Pyraloidea Lepidoptera Chewer RLK >994 >241 50 ca. 3,500 Arctiidae Lepidoptera Chewer RLK >432 >74 74 ca. 600 Psocoptera Psocoptera Epiphyte-grazer PC 212 66 2 112 Orthoptera Orthoptera Various >2,500 130 0 - IBISCA - Panama 61 biodiversity. In particular, international collaboration Acknowledgements based on an integrated experimental design allows the investigation of crucial ecological questions in IBISCA-Panama is an initiative of Pro-Natura International, Océan unprecedented ways. Since the Panamanian experience, Vert, l’Université Blaise Pascal, la Universidad de Panama and the Smithsonian Tropical Research Institute (STRI), with core funding IBISCA participants have decided that the IBISCA from SolVin-Solvay SA, STRI, the United Nations Environment model should be expanded to sites in other geographic Programme, the Smithsonian Institution (Walcott fund), the regions, so as to address more effectively a range of European Science Foundation and the Global Canopy Programme. issues of pressing importance related to biodiversity. The Canopy Crane team (José Herrera, Edwin Andrade) and Joe Wright at STRI; the tree climbers from “Les ACCRO-branchés” IBISCA-Queensland began in eastern Australia in (Nouï Baiben, Stéphane Bechet & Julien Belleguic) as well as April 2006 at 20 sites (now referred to as “standard” Thierry Aubert and Kevin Jordan; “Océan Vert” (Gilles Ebersolt, 20 x 20 m IBISCA-plots). Four plots were each Dany Cleyet-Marrel & Laurent Pyot); and Patrick Basset & Eric located at five different altitudes between 300 m and Bauhaus helped with logistics in the field. Anovel Barba, Rosario 1100 m within the continuum of forest in Lamington Cabrera, Isella Díaz, Mario Gonzalez, Jorge Herrera, Betzi Perez and Oldemar Valdez helped with initial sorting of the arthropod National Park. Three main field periods are scheduled: and plant material. Lawrie Springate kindly checked botanical September 2006, March 2007 and January 2008. Two nomenclature. Roger Le Guen and Laurent Pyot provided free use of these have been completed to date. IBISCA-Santo of their photos. Isabelle Bachy helped to edit the figures and plates. took place in October-December 2006 in Vanuatu, as IBISCA is greatly honoured to be under the patronage of Professor E.O. Wilson of Harvard University. We dedicate this article to part of the “Forests, Mountains and Rivers” component the late Professor Joachim Adis, tropical ecologist, entomologist, of the Santo 2006 Global Biodiversity Survey. As with colleague and friend. IBISCA-Queensland, IBISCA-Santo will examine distribution patterns of arthropods and plants along an altitudinal gradient (100 m to 1200 m in the case References of IBISCA-Santo). Two other IBISCA projects are currently in preparation: one in a temperate deciduous ADIS, J., BASSET, Y., FLOREN, HAMMOND, P.M. & LINSENMAIR, forest (Forêt de la Comté d’Auvergne, France); and K.E., 1998. Canopy fogging of an overstorey tree - the second in a coastal tropical dry forest in Northern recommendations for standardization. Ecotropica, 4: 93-97. Mozambique (see www.ibisca.net for up-to-date ADIS, J. & SCHUBART, H.O.R., 1984. Ecological research on information related to IBISCA programmes). arthropods in Central Amazonian forest ecosystems with We urgently need these projects, and many more recommendation for study procedures. In: COOLEY, J. & like them, to understand how arthropod biodiversity GOLLEY, F.B. (eds.). Trends in Ecological Research for the originated, how it is currently distributed and 1980’s. Plenum Press, New York, pp. 111-144. maintained in temperate and tropical forests, and to ANDRÉ, H.M., DUCARME, X. & LEBRUN, P., 2002. Soil address biodiversity issues in an ecological context. biodiversity: myth, reality or conning? Oikos, 96: 3-24. Politicians and funding agencies should recognize that ANDRÉ, H.M., LEBRUN, P. & NOTI, M.-I., 1992. Biodiversity biodiversity research is no less technically challenging in Africa: a plea for more data. Journal of African Zoology, than is gene discovery, astrophysics, or other large- 106: 3-15. scale scientific endeavours. The costs of protecting ecosystems under study and of developing research BARKER, M. & STANDRIDGE, N., 2002. Ropes as mechanisms for canopy access. In: MITCHELL, A.W., SECOY, K. & JACKSON, infrastructure must be judged to be as justified as T. (eds.). The Global Canopy Handbook. Global Canopy the development and massive use of ‘cutting edge’ Programme, Oxford, pp. 13-23. molecular tools. Knowledge to be gained from understanding the evolution and functioning of a BARRIOS, H. & MEDIANERO, E., 1999. Riqueza de insectos tropical forest or coral reef is every much as powerful formadores de agallas en nueve zonas de vida de la República de Panamá. Scientia (Panamá), 14: 39-58. as that associated with the sequencing of a genome. The intense spatial focus and collegial synergies of BASSET, Y., 2001. Communities of insect herbivores projects such as IBISCA cannot be underestimated. We foraging on mature trees vs. seedlings of Pourouma bicolor are calling for institutional mechanisms that will better (Cecropiaceae) in Panama. Oecologia, 129: 253-260. facilitate and support such teamwork in the future. We BASSET, Y., ABERLENC, H.-P., SPRINGATE, N.D. & DELVARE, G., know the key questions. We have the methods to answer 1997. A review of methods for collecting arthropods in tree them. We need wider appreciation of the challenges canopies. In: STORK, N.E., ADIS, J.A. & DIDHAM, R.K. (eds.). faced by biodiversity research and the resources to face Canopy Arthropods. Chapman & Hall, London, pp. 27-52. them. We trust that the former will lead to the latter. BASSET, Y., HORLYCK, V. & WRIGHT, S.J. (eds.), 2003b. 62 Yves BASSET et al.

Studying forest canopies from above: The International pp. 61-64. Canopy Crane Network. Smithsonian Tropical Research CURLETTI, G., 2005. Notes about the genus Agrilus Curtis, Institute and UNEP, Panama. 1825 in Panama. Lambillionea, 105: 545-571. BASSET, Y., NOVOTNY, V., MILLER, S.E. & KITCHING, R.L. (eds.), CURLETTI, G., ABERLENC, H.-P. & BASSET, Y., 2006. Progetto 2003a. Arthropods of Tropical Forests. Spatio-temporal IBISCA in Panama: considerazioni sul genere Agrilus Curtis, Dynamics and Resource Use in the Canopy. Cambridge 1825. Rivista Piemontese di Storia Naturale, 27: 339-348. University Press, Cambridge. DIDHAM, R.K. & FAGAN, L.L., 2003. Project IBISCA – BASSET, Y., NOVOTNY, V., MILLER, S. E., WEIBLEN, G. D., MISSA, Investigating the biodiversity of soil and canopy arthropods. O. & STEWART, A.J.A., 2004. Conservation and biological The Weta, 26: 1-6. monitoring of tropical forests: the role of parataxonomists. Journal of Applied Ecology, 41: 163-174. DIDHAM, R.K. & FAGAN, L.L., 2004. Forest canopies. In: BURLEY, J., EVANS, J. & YOUNGQUIST, J. (eds.). Encyclopaedia BASSET, Y. & SAMUELSON, G.A., 1996. Ecological of Forest Sciences. Academic Press, Elsevier Science, characteristics of an arboreal community of Chrysomelidae London, pp. 68-80. in Papua New Guinea. In: JOLIVET, P.H.A. & COX, M.L. (eds.). Chrysomelidae Biology. Volume 2. Ecological Studies. SPB ERWIN, T.L., 1982. Tropical forests: their richness in Academic Publishing, Amsterdam, pp. 243-262. Coleoptera and other arthropod species. The Coleopterists Bulletin, 36: 74-75. BRADBURY, J., 2003. Surveying Panama’s treetops. Frontiers in Ecology and Environment, 9: 457. ERWIN, T.L. & SCOTT, J.C., 1980. Seasonal and size patterns, trophic structure and richness of Coleoptera in the tropical CASTAÑO-MENESES, G., BASSET, Y., WINCHESTER, N. & arboreal ecosystem: the fauna of the tree Luehea seemannii BARRIOS, H., 2006. Colembolos (Hexapoda: Collembola) Triana and Planch in the Canal Zone of Panama. The del dosel en la selva tropical de San Lorenzo, Provincia de Coleopterists Bulletin, 34: 305-322. Colon, Panama. Entomologia mexicana, 5: 486-490. EWERS, R.M., THORPE, S. & DIDHAM, R.K., 2007. Synergistic CHARLES, E. & BASSET, Y., 2005. Stratification of leaf interactions between edge and area effects in a heavily assemblages (Coleoptera: Chrysomelidae) in two forest types fragmented landscape. Ecology, 88: 96-106. in Panama. Journal of Tropical Ecology, 21: 329-336. FAGAN, L.L., DIDHAM, R.K., WINCHESTER, N.N., BEHAN- CHAVE, J., CONDIT, R., AGUILAR, S., HERNANDEZ, A., LAO, S. PELLETIER, V., CLAYTON, M., LINDQUIST, E. & RING, R.A., & PÉREZ, R., 2004. Error propagation and scaling for tropical 2006. An experimental assessment of biodiversity and forest biomass estimates. Philosophical Transactions of the species turnover in terrestrial vs canopy leaf litter. Oecologia, Royal Society, Series B, 359: 409-420. 147: 335-347. COLWELL, R.K., 2004. EstimateS, Version 7: Statistical FERRER, J. & ØDEGAARD, F., 2005. New species of darkling Estimation of Species Richness and Shared Species from beetles from Central America with systematic notes Samples. User’s Guide and application published at: (Coleoptera: Tenebrionidae). Annales Zoologici, 55: 633- http://purl.oclc.org/estimates. 661. CONDIT, R., AGUILAR, S., HERNANDEZ, A., PÉREZ, R., LAO, S., FISHER, B.L., 1998. Ant diversity patterns along an elevational ANGEHR, G., HUBBELL, S.P. & FOSTER, R.B., 2004. Tropical gradient in the réserve spéciale d’Anjanaharibe-Sud and on forest dynamics across a rainfall gradient and the impact the western Masoala peninsula, Madagascar. Fieldiana of an El Niño dry season. Journal of Tropical Ecology, 20: Zoology (n.s.), 90: 39-67. 51–72. FROST, S.W., 1957. The Pennsylvania light trap. Journal of CONDIT, R. & PÉREZ, R. 2002. Tree Atlas of the Panama Canal Economic Entomology, 50: 287-292. Watershed. Center for Tropical Forest Science, Panama. http://ctfs.si.edu/webatlas/maintreeatlas.html. GERARDO, N.M., MUELLER, U.G., PRICE, S.L. & CURRIE, C.R., 2004. Exploiting a mutualism: parasite specialization CONDIT, R., WATTS, K., BOHLMAN, S.A., PÉREZ, R., FOSTER, on cultivars within the fungus-growing ant symbiosis. R.B. & HUBBELL, S.P., 2000. Quantifying the deciduousness Proceedings of the Royal Society of London, Series B, 271: of tropical forest canopies under varying climates. Journal of 1791-1798. Vegetation Science, 11: 649-658. GODFRAY, H.C., LEWIS, O.T. & MEMMOTT, J., 1999. Studying CORBARA, B., BASSET, Y. & BARRIOS, H. 2006. IBISCA: insect diversity in the tropics. Philosophical Transactions of a large-scale study of arthropod mega-diversity in a the Royal Society, Series B, 354: 1811-1824. neotropical rainforest. In: SEGERS, H., DESMET, P. & BAUS, E. (eds.). Tropical Biodiversity: Science, Data, Conservation. GRASSLE, J.F. & SMITH, W., 1976. A similarity measure Proceedings of the 3rd GBIF Science Symposium, Brussels, sensitive to the contribution of rare species and its use in 18-19 April 2005. Belgian Biodiversity Platform, Brussels, investigation of variation in marine benthic communities. IBISCA - Panama 63

Oecologia, 25: 13-22. KITCHING, R.L., ORR, A.G., THALIB, L., MITCHELL, H., HOPKINS, M.S. & GRAHAM, A.W., 2000. Moth assemblages as indicators HALLÉ, F. (ed.), 2000. Biologie d’une Canopée de Forêt of environmental quality in remnants of upland Australian Tropicale - IV. Rapport de la Mission du Radeau des Cimes à rain forest. Journal of Applied Ecology, 37: 284-297. la Makandé, Forêt des Abeilles, Gabon, Janvier-Mars 1999. Pro-Natura International & Opération Canopée, Paris. LAVELLE, P., LATTAUD, C., TRIGO, D. & BAROIS, I., 1995. Mutualism and biodiversity in soils. Plant & Soil, 170: 23- HAMMOND, P.M., 1990. Insect abundance and diversity in 33. the Dumoga-Bone National Park, N. Sulawesi, with special reference to the beetle fauna of lowland rain forest in the LAWTON, J.H., BIGNELL, D.E., BOLTON, B., BLOEMERS, G.F., Toraut region. In: KNIGHT, W.J. & HOLLOWAY, J.D. (eds.). EGGLETON, P., HAMMOND, P.M., HODDA, M., HOLT, R.D., Insects and the Rain Forests of South East Asia (Wallacea). LARSEN, T.B., MAWDSLEY, N.A., STORK, N.E., SRIVASTAVA, The Royal Entomological Society of London, London, pp. D.S. & WATT, A.D., 1998. Biodiversity inventories, indicator 197-254. taxa and effects of habitat modification in a tropical forest. HAMMOND, P.M., 1995. Magnitude and distribution of Nature, 391: 72-76. biodiversity. In: HEYWOOD, V.T. & WATSON, R.T. (eds.). LEPONCE, M., THEUNIS, L., DELABIE, J.H.C. & ROISIN, Y., 2004. Global Biodiversity Assessment. Cambridge University Scale dependence of diversity measures in a leaf-litter ant Press, Cambridge, pp. 113-138. assemblage. Ecography, 27: 253-267. HAMMOND, P.M., STORK, N.E. & BRENDELL, M.J.D., 1997. LINDO, Z. & WINCHESTER, N.N., 2006. A comparison of Tree-crown beetles in context: a comparison of canopy and microarthropod assemblages with emphasis on oribatid mites other ecotone assemblages in a lowland tropical forest in in canopy suspended soils and forest floors associated with Sulawesi. In: STORK, N.E., ADIS, J.A. & DIDHAM, R.K. (eds.). ancient western redcedar trees. Pedobiologia, 50: 31-41. Canopy Arthropods. Chapman & Hall, London, pp. 184-223. LINSENMAIR, K.E., DAVIS, A.J., FIALA, B. & SPEIGHT, M.R. HARDY, O.J., 2007. BiodivR 1.0. A program to compute (eds.), 2001. Tropical Forest Canopies: Ecology and statistically unbiased indices of species diversity within Management. Kluwer Academic Publishers, Dordrecht, The samples and species similarity between samples using Netherlands. rarefaction principles. Published at: http://www.ulb.ac.be/ sciences/ecoevol/biodivr.html. LONGINO, J.T., 2004. Arthropods of the tropical canopy (and elsewhere). Ecology, 85: 1170-1171. HOOPER, D.U., BIGNELL, D.E., BROWN, V.K., BRUSSAARD, L., DANGERFIELD, J.M., WALL, D.H., WARDLE, D.A., COLEMAN, LOWMAN, M.D. & NADKARNI, N.M. (eds.), 1995. Forest D.C., GILLER, K.E., LAVELLE, P., VAN DER PUTTEN, W.H., DE Canopies. Academic Press, San Diego. RUITER, P.C., RUSEK, J., SILVER, W.L., TIEDJE, J.M. & WOLTERS, MAY, R.M., 1990. How many species? Philosophical V., 2000. Interactions between aboveground and belowground Transactions of the Royal Society, Series B, 330: 293-304. biodiversity in terrestrial ecosystems: patterns, mechanisms, and feedbacks. BioScience, 50: 1049-1061. MANN, C.C., 1991. Extinction: are ecologists crying wolf? Science, 253: 736-738. HUBBELL, S.P., 2001. The Unified Neutral Theory of Biodiversity and Biogeography. Princeton University Press, MEDIANERO, E., CASTAÑO-MENESES, G., TISHECHKIN, A., BASSET, Princeton. Y., BARRIOS, H, ØDEGAARD, F., CLINE, A.R. & BAIL, J., 2007. Influence of local illumination and plant composition on the KWESKIN, M.P., 2004. Jigging in the fungus-growing ant spatial and seasonal distribution of litter-dwelling arthropods Cyphomyrmex costatus: a response to collembolan garden in a tropical forest. Pedobiologia, 51: 131-145. invaders? Insectes Sociaux, 51: 158–162. MEDIANERO, E., VALDERRAMA, A. & BARRIOS, H., 2003. KIRKENDALL, L.R. & ØDEGAARD, F. 2007. Ongoing invasions of Diversidad de insectos minadores de hojas y formadores de old-growth tropical forests: establishment of three incestuous agallas en el dosel y sotobosque del bosque tropical. Acta beetle species in southern Central America (Curculionidae: Zoológica Mexicana (n.s.), 89: 153-168. Scolytinae). Zootaxa, in press. MILLER, S.E., 2000. Taxonomy for understanding biodiversity KITCHING, R.L., 2006. Crafting the pieces of the biodiversity - the strengths of the BioNet International model. In: JONES, jigsaw. Science, 313: 1055-1057. T. & GALLAGHER, S. (eds.). Proceedings of the Second KITCHING, R.L., BASSET, Y., OZANNE, C. & WINCHESTER, N., BioNet-International Global Workshop (BIGW2). BioNet 2002. Canopy knockdown techniques for sampling canopy International, Egham, Surrey, England, pp. 13-18. arthropod fauna. In: MITCHELL, A.W., SECOY, K. & JACKSON, T. MITCHELL, A.W., SECOY, K. & JACKSON, T. (eds.), 2002. Global (eds.). The Global Canopy Handbook. Techniques of Access Canopy Handbook. Techniques of Access and Study in the and Study in the Forest Roof. Global Canopy Programme, Forest Roof. Global Canopy Programme, Oxford. Oxford, pp. 134-139. 64 Yves BASSET et al.

MUIRHEAD-THOMSON, R.C., 1991. Trap Responses of Flying PYKE, C.R., CONDIT, R., AGUILAR, S. & LAO, S., 2001. Floristic Insects. Academic Press, London. composition across a climatic gradient in a neotropical lowland forest. Journal of Vegetation Science, 12: 553–566. NOVOTNY, V., BASSET, Y., MILLER, S.E., WEIBLEN, G.D., BREMER, B., CIZEK, L. & DROZD, P., 2002. Low host specificity QUINTERO, D., & AIELLO, A. (eds.), 1992. Insects of Panama of herbivorous insects in a tropical forest. Nature, 416: 841- and Mesoamerica: Selected studies. Oxford University Press, 844. Oxford and New York.

ØDEGAARD, F., 2000. How many species of arthropods? RIBEIRO, S.P. & BASSET, Y., 2007. Gall-forming and free- Erwin’s estimate revised. Biological Journal of the Linnean feeding herbivory along vertical gradients in a lowland Society, 71: 583-597. tropical rainforest: the importance of leaf sclerophylly. Ecography, in press. ØDEGAARD, F., 2003. Taxonomic composition and host specificity of phytophagous beetles in a dry forest in Panama. ROISIN, Y., DEJEAN A., CORBARA B., ORIVEL J., SAMANIEGO, M. In: Basset, Y., Novotny, V., Miller, S.E. & Kitching, R.L. & LEPONCE, M., 2006. Vertical stratification of the termite (eds.). Arthropods of tropical forests: Spatio-temporal assemblage in a neotropical rainforest. Oecologia, 149: 301- dynamics and resource use in the canopy. Cambridge 311. University Press, Cambridge, pp. 220-236. ROSLIN, T., 2003. Not so quiet on the high frontier. Trends in ØDEGAARD, F., 2004. Species richness of phytophagous Ecology and Evolution, 18: 376-379. beetles in the tropical tree Brosimum utile (Moraceae): the SAGERS, C.L. & COLEY, P.D., 1995. Benefits and costs of effects of sampling strategy and the problem of tourists. defense in a Neotropical shrub. Ecology, 76: 1835-1843. Ecological Entomology, 29: 76-88. SALA, O.E., CHAPIN, F.S., III, ARMESTO, J.J., BERLOW, E., ØDEGAARD, F., 2006. Host specificity, alpha- and beta- BLOOMFIELD, J., DIRZO, R., HUBER-SANWALD, E., HUENNEKE, diversity of phytophagous beetles in two tropical forests in L.F., JACKSON, R.B., KINZIG, A., LEEMANS, R., LODGE, D.M., Panama. Biodiversity and Conservation, 15: 83-105. MOONEY, H.A., OESTERHELD, M., LEROY POFF, N., SYRKES, M. ØDEGAARD, F., DISERUD, O.H. & ØSTBYE, K., 2005. The T., WALKER, B.H., WALKER, M. & WALL, D.H., 2000. Global importance of plant relatedness for host utilization among biodiversity scenarios for the year 2100. Nature, 287: 1770- phytophagous insects. Ecology Letters, 8: 612-617. 1774.

ØDEGAARD, F. & FRAME, D., 2007. Generalist flowers and SCHMIDL, J. & CORBARA, B., 2005. IBISCA – Artenvielfalt der phytophagous beetles in two tropical canopy trees: resources Boden- und Baumkronen-Arthropoden in einem tropischen for multitudes. Taxon, 56: 696-706. Regenwald (San Lorenzo NP, Panama). Entomologische Zeitschrift, 115: 104-108. OSLER, G.H.R. & BEATTIE, A.J., 2001. Contribution of oribatid and mesostigmatid soil mites in ecologically based estimates SCHNITZER, S.A., 2005. A mechanistic explanation for global of global species richness. Austral Ecology, 26: 70-79. patterns of liana abundance and distribution. The American Naturalist, 166: 262-276. OZANNE, C.M.P., ANHUF, D., BOULTER, S.L., KELLER, M., KITCHING, R.L., KÖRNER, C., MEINZER, F.C., MITCHELL, A.W., SCHOWALTER, T.D. & GANIO, L.M., 2003. Diel, seasonal and NAKASHIZUKA, T., SILVA DIAS, P.L., STORK, N.E., WRIGHT, S.J. disturbance-induced variation in invertebrate assemblages. & YOSHIMURA, M., 2003. Biodiversity meets the atmosphere: In: BASSET, Y., NOVOTNY, V., MILLER, S.E. & KITCHING, a global view of forest canopies. Science, 301: 183-186. R.L. (eds.). Arthropods of tropical forests: Spatio-temporal dynamics and resource use in the canopy. Cambridge PALACIOS-VARGAS, J.G., CASTAÑO-MENESES, G. & GÓMEZ- University Press, Cambridge, pp. 315-328. ANAYA, J.A., 1998. Collembola from the canopy of a Mexican tropical decidous forest. Pan-Pacific Entomologist, SOUTHWOOD, T.R.E., 1978. Ecological Methods, with 74: 47-54. Particular Reference to the Study of Insect Populations. 2nd ed. Chapman and Hall, London. PARKER, G.G., SMITH, A.P. & HOGAN, K.P., 1992. Access to the upper forest canopy with a large tower crane. BioScience, SPRINGATE, N.D. & BASSET, Y., 2004. IBISCA 2003-2005, 42: 664-670. Panama: progress report. Bulletin of the British Ecological Society, 35: 21-23. PENNISI, E., 2005. Sky-high experiments. Science, 309: 1314- 1315. STORK, N.E., 1988. Insect diversity: facts, fiction and speculation. Biological Journal of the Linnean Society, 35: POKON, R., NOVOTNY, V. & SAMUELSON, G.A., 2005. 321-337. Host specialization and species richness of root-feeding chrysomelid larvae (Chrysomelidae, Coleoptera) in a New STORK, N.E., 1991. The composition of the arthropod fauna Guinea rain forest. Journal of Tropical Ecology, 21: 595- of Bornean lowland rain forest trees. Journal of Tropical 604. Ecology, 7: 161-180. IBISCA - Panama 65

STORK, N.E., 1996. Tropical forest dynamics: the faunal WEAVER, P.L. & BAUER, G.P., 2004. The San Lorenzo Protected components. In: Edwards, D.S., Booth, W.E. & Choy, S.C. Area: A Summary of Cultural and Natural Resources. (eds.). Tropical Rainforest Research - Current Issues. Kluwer International Institute of Tropical Forestry, San Juan, PR. Academic Publishers, Amsterdam, pp. 1-20. WELDEN, C.W., HEWETT, S.W., HUBBELL, S.P. & FOSTER, R.B., STORK, N.E., ADIS, J.A. & DIDHAM, R.K. (eds.), 1997. Canopy 1991. Sapling survival, growth and recruitment: relationship Arthropods. Chapman & Hall, London. to canopy height in a Neotropical forest. Ecology, 72: 35-50.

STORK, N.E., & GRIMBACHER, P.S., 2006. Beetle assemblages WINCHESTER, N.N. & BEHAN-PELLETIER, V., 2003. Fauna from an Australian tropical rainforest show that the canopy of suspended soils in an Ongokea gore tree in Gabon. In: and the ground strata contribute equally to biodiversity. Basset, Y., Novotny, V., Miller, S.E. & Kitching, R.L. (eds.). Proceedings of the Royal Society of London, Series B, 273: Arthropods of Tropical Forests. Spatio-temporal Dynamics 1969-1975. and Resource Use in the Canopy. Cambridge University Press, Cambridge, pp. 102-109. TOWNES, H., 1972. A light-weight Malaise trap. Entomological News, 83: 239-247. WRIGHT, S.J., HORLYCK, V., BASSET, Y., BARRIOS, H., BETHANCOURT, A., BOHLMAN, S.A., GILBERT, G.S., GOLDSTEIN, VEECH, J.A., SUMMERVILLE, K.S., CRIST, T.O. & GERING, J.C., G., GRAHAM, E.A., KITAJIMA, K., LERDAU, M.T., MEINZER, 2002. The additive partitioning of species diversity: recent F.C., ØDEGAARD, F., REYNOLDS D.R. , ROUBIK, D.W., SAKAI, revival of an old idea. Oikos, 99: 3-9. S., SAMANIEGO, M., SPARKS, J.P., VAN BAEL, S., WINTER K. & VILLELSEN, P., SCHULTZ, T.R. & BOOMSMA, J.J., 2002. ZOTZ, G., 2003. Tropical Canopy Biology Program, Republic Identifying the transition between single and multiple mating of Panama. In: BASSET, Y., HORLYCK, V. & WRIGHT S.J. (eds.). in attine ants. Proceedings of the Royal Society of London, Studying Forest Canopies from Above: The International Series B, 269: 1541-1548. Canopy Crane Network. Smithsonian Tropical Research Institute and UNEP, Panama City, pp. 138-158. Available WALTER, D.E., & BEHAN-PELLETIER, V., 1999. Mites in forest on-line at: http://www.stri.org/english/research/facilities/ canopies: Filling the size distribution shortfall? Annual terrestrial/cranes/canopy_crane_network.php. Review of Entomology, 44: 1-19. ZAYED, A., ROUBIK, D.W. & PACKER, L., 2003. Use of diploid WALTER, D.E., SEEMAN, O., RODGERS, D. & KITCHING, R.L., male frequency data as an indicator of pollinator decline. 1998. Mites in the mist: how unique is a rainforest canopy- Proceedings of the Royal Society of London, Series B, 271: knockdown fauna? Australian Journal of Ecology, 23: 501- S9-S12. 508. ZOTZ, G., 2004. How prevalent is crassulacean acid WCISLO, W.T., ARNESON, L., ROESCH, K., GONZALEZ, V., SMITH, metabolism among vascular epiphyes? Oecologia, 138: 184- A.M. & FERNANDEZ-MARIN, H., 2004. The evolution of 192. nocturnal behaviour in sweat bees, Megalopta genalis and M. ecuadoria (Hymenoptera: Halictidae): an escape from competitors and enemies? Biological Journal of the Linnean Society, 83: 377-387. 66 Yves BASSET et al.

1Smithsonian Tropical Research Institute, Panama City, 18Crop and Food Research, Lincoln, New Zealand, Republic of Panama, [email protected] (YB, author [email protected] for correspondence), [email protected] (AH), 19Universität Würzburg, Würzburg , Germany, [email protected] (MS) & [email protected] (DWR) [email protected] 2Université Blaise Pascal, Clermont-Ferrand, France, 20University of Montpellier, Montpellier, [email protected] France, [email protected] (DF) and 3Universidad de Panamá, Panama City, Republic of [email protected] (FH) Panama, [email protected] 21Université Libre de Bruxelles, Brussels, Belgium, 4Muséum d’histoire naturelle de la Ville de Genève, [email protected] (OJH) & [email protected] (YR) Genève, Switzerland, [email protected] 22Universidade Estadual de Campinas, São Paulo, 5Royal Belgian Institute of Natural Sciences, Brussels, Brazil, [email protected] Belgium, [email protected] 23University of Oxford, Oxford, U. K., 6CIRAD, Montpellier, France,henri- [email protected] [email protected] 24The New Guinea Binatang Research Center, Madang, 7University of Erlangen, Erlangen, Germany, Papua New Guinea, [email protected] [email protected] (JB) & 25Museo Nacional de Ciencias Naturales, Madrid, [email protected] (JS) Spain, [email protected] 8Griffith University, Brisbane, Australia, 26University of York, York, U. K, [email protected] [email protected] (DB) & [email protected] (RLK) 27Global Canopy Programme, Oxford, U. K., andrew.m [email protected] 9University of Bristol, Bristol, U. K., [email protected] 28Norwegian Institute for Nature Research, Trondheim, Norway, [email protected] 10Universidad Nacional Autónoma de México, México, [email protected] 29Conservation International (TEAM), Brazil, [email protected] 11Biological Centre, Czech Academy of Sciences, Ceske Budejovice, Czech Republic, 30Roehampton University, London, U. K., [email protected] (LC) [email protected] & [email protected] (VN) 31Pro Natura International, Paris, France, 12Movimiento de Investigaciones Biológicas [email protected] de Panamá, Panama City, Republic of 32Imperial College, London, U. K., Panama, [email protected] (AC) & [email protected] [email protected] (MR) 33Universidade Federal de Ouro Preto, Ouro Preto, 13 Museo Civico di Storia Naturale, Carmagnola, Italy, Brazil, [email protected] [email protected] 34University of Helsinki, Helsinki, Finland, 14 Centro de Pesquisas do Cacau, Itabuna, Brazil, [email protected] [email protected] 35Natural History Museum of Denmark, Copenhagen, 15 EDB, UMR-CNRS 5174, University Toulouse Denmark, [email protected] III, Toulouse, France, [email protected] (AD) & [email protected] (JO) 36Lousiana State Arthropod Museum, Baton Rouge, USA, [email protected] 16University of Canterbury, Christchurch, New Zealand, [email protected] 37Biotrac SPRL, Brussels, Belgium, [email protected] 17Ministère de la Région Wallonne, Gembloux, Belgium, [email protected] 38University of Victoria, Victoria, Canada, [email protected] IBISCA - Panama 67

Appendix 1 — Frequency of 165 tree species occurring at the 12 IBISCA sites, listed by species. Nomenclature follows CONDIT and PÉREZ (2002), further checked against resources at http://mobot.mobot.org/W3T/Search/vast.html, http://www.ipni.org/index.html and the Royal Botanic Garden, Edinburgh, by Lawrie Springate.

Code Plant Species Family B1 B2 C1 C2 C3 F1 F2 F3 I1 R1 R2 R3

PIT1BA Abarema barbouriana (Standl.) Barneby & J.W. Grimes Fabaceae:Mimos. ------1 - ACALDI Acalypha diversifolia Jacq. Euphorbiaceae ------1 - - AEGIPA Aegiphila panamensis Moldenke Verbenaceae ------1 ALCHLA Alchornea latifolia Pax & K. Hoffm. in Engl. Euphorbiaceae - - 1 ------AMAICO Amaioua corymbosa Kunth Rubiaceae ------1 - - - - ANACEX Anacardium excelsum (Bert. & Balb. ex Kunth) Skeels Anacardiaceae ------1 - - ANDIIN Andira inermis (W. Wright) DC. Fabaceae:Papil. - - 2 - 1 - - - 1 - - - APEIME Apeiba membranacea Spruce ex Benth. Tiliaceae - 1 1 - 1 - - - 2 - - - APEITI Apeiba tibourbou Aubl. Tiliaceae - - 1 ------ARDIBA Ardisia bartlettii Lundell Myrsinaceae ------1 - 1 - ASPICR Aspidosperma cruentum Benth. ex Müll.Arg. Apocynaceae - - - - 6 1 1 5 1 - 4 1 AST1ST Astrocaryum standleyanum L.H. Bailey Arecaceae ------4 - - BACTC1 Bactris coloniata L. H. Bailey Arecaceae - 2 1 - 2 2 - - 2 1 - 3 BACTC2 Bactris coloradonis L.H. Bailey Arecaceae - 1 - 2 1 - 1 4 1 - - - BACTMA Bactris major Jacq. Arecaceae - - - - - 1 - - - - - 1 BACTPA Bactris panamensis de Nevers & Grayum Arecaceae - - - 1 ------BEILPE Beilschmiedia pendula (Sw.) Hemsl. Lauraceae ------1 BROSGU Brosimum guianense (Aubl.) Huber Moraceae 1 1 - 3 - - - 3 3 - 3 - BROSUT Brosimum utile var. utile (Kunth) Pittier Moraceae 6 21 1 2 5 7 2 5 8 - 4 4 BYRSCR Byrsonima crassifolia (L.) Kunth Malpighiaceae 1 ------CALOLO Calophyllum longifolium Willd. Clusiaceae 4 - - 1 - 1 1 1 2 - 2 - CARAGU Carapa guianensis Aubl. Meliaceae 2 2 3 1 1 4 6 4 1 6 - 3 CASECO Casearia commersoniana Cambess. Flacourtiaceae - 5 - - 1 1 - 1 - - 1 1 CASESY Casearia sylvestris Sw. Flacourtiaceae ------1 1 - - CASSEL Cassipourea elliptica (Sw.) Poir. Rhizophoraceae - - - 1 - - - 1 - - - - CECRIN Cecropia insignis Liebm. Cecropiaceae - - 2 ------1 CECRPE Cecropia peltata L. Cecropiaceae - - 1 ------CESPMA Cespedezia spathulata Planch. Ochnaceae 1 - 2 1 1 1 - 1 - - 1 - CESTME Cestrum megalophyllum Dunal Solanaceae ------1 - - CHA1TE Chamaedorea tepejilote Liebm. ex Mart. Arecaceae ------1 - - CHR2AR Chrysophyllum argenteum Jacq. Sapotaceae 2 - 1 - 1 - - - 1 - - - COCCPA Coccoloba padiformis Meisn. in A. DC. Polygonaceae ------1 - - - COURPA Couratari guianensis Aubl. Lecythidaceae ------2 2 - - - - CREMPA Cremastosperma panamense Maas Annonaceae - 1 2 - - 1 1 1 - - - 2 CUPASC Cupania scrobiculata Rich. Sapindaceae - 3 1 - - 1 - - 1 - 5 - CYM1LA Cymbopetalum lanugipetalum Schery Annonaceae - - 1 - 1 - 1 - 3 - - - DENDAR Dendropanax arboreus (L.) Dec. & Planch Araliaceae 3 - 2 - 4 3 3 11 4 - 2 2 DES2PA Desmopsis panamensis (B.L. Rob.) Saff. Annonaceae - - - - 1 - - - 1 8 - - DIO2AR Diospyros artanthifolia Mart. ex Miq. Ebenaceae - - 1 ------1 - DISCGU Discophora guianensis Miers Icacinaceae - - - - 2 - - 4 - - 1 - DUSS1 Dussia sp.1 Fabaceae:Papil. 1 - 1 1 1 ------ERY1CO Erythrina costaricensis Micheli Fabaceae:Papil. 2 ------ERY2MA Erythroxylum macrophyllum Cav. Erythroxylaceae ------1 EUGECO Eugenia coloradoensis Standl. Myrtaceae ------1 - 1 - - 1 EUGESP Eugenia sp. Myrtaceae ------1 - - - EUTEPR Euterpe precatoria Mart. Arecaceae ------1 - 2 - FARAOC Faramea occidentalis (L.) A. Rich. Rubiaceae ------10 - - FICUOB Ficus obtusifolia Kunth in Humb., Bonpl. & Kunth Moraceae ------1 - - - - - FICUTO Ficus tonduzii Standl. Moraceae - - 1 ------GAR2MA Garcinia madruno (H. B. K.) Hammel Clusiaceae - - - 1 - - 1 - 1 - 4 - GEONCO Geonoma congesta H. Wendl. ex Spruce Arecaceae 1 21 7 8 8 16 31 14 4 - 1 24 GEONCU Geonoma cuneata H. Wendl. ex Spruce Arecaceae - - 4 4 6 12 13 3 2 - - 2 GUARGU Guarea guidonia (L.) Sleumer Meliaceae 1 ------1 - - GUAR2 Guarea sp. ‘sherman’ Meliaceae - - - 1 - - - - - 2 - - GUATAM Guatteria amplifolia Triana & Planch. Annonaceae - - 1 - 1 1 - 2 - - 1 - GUATDU Guatteria dumetorum R.E. Fr. Annonaceae 2 - 2 - - 3 1 2 - - 1 2 GUETFO Guettarda foliacea Standl. Rubiaceae ------1 - - 68 Yves BASSET et al.

HEISAC Heisteria acuminata (Humb. & Bonpl.) Engl. Olacaceae 1 - - 1 3 - 1 1 - - 1 - HENRSU Henriettea succosa (Aubl.) DC. Melastomataceae ------1 - HIRTRA Hirtella racemosa Lam. Chrysobalanaceae ------2 1 - - - HIRTTR Hirtella triandra Sw. Chrysobalanaceae - - 1 ------5 - - HUMIDI Humiriastrum diguense (Cuatrec.) Cuatrec. Humiriaceae ------2 1 - - - - HYERAL Hyeronima alchorneoides Allemão Euphorbiaceae 1 ------HYM1ME Hymenolobium mesoamericanum H.C. Lima Fabaceae:Papil. - - 1 - - 1 ------INGACO Inga cocleensis Pittier Fabaceae:Mimos. - - 1 - - - - - 1 - - - INGAGO Inga goldmanii Pittier Fabaceae:Mimos. 1 - 1 ------INGAMA Inga marginata Willd. Fabaceae:Mimos. 1 ------INGAM2 Inga multijuga Benth. Fabaceae:Mimos. 2 - - - - 1 - - 1 - - - INGAPE Inga pezizifera Benth. Fabaceae:Mimos. 3 - 3 1 - 2 1 - 2 - - - INGASE Inga sertulifera DC. Fabaceae:Mimos. 11 - 2 ------JAC1CO Jacaranda copaia (Aubl.) D. Don Bignoniaceae - 1 - - 1 - - 1 - - - - LACIAG Lacistema aggregatum (P.J. Bergius) Rusby Flacourtiaceae 1 - 1 - 3 - - 2 1 3 3 1 LACMPA Lacmellea panamensis (Woodson) Markgr. Apocynaceae 7 - 1 - - - - 1 - - - - LAETPR Laetia procera (Poepp.) Eichler Flacourtiaceae - 2 ------LICAHY Licania hypoleuca Benth. Chrysobalanaceae - 2 1 1 - 1 ------LONCLA Lonchocarpus latifolius (Willd.) DC. Fabaceae:Papil. - - 2 ------1 - LOZAPI Lozania pittieri (S.F. Blake) L.B. Sm. Flacourtiaceae 5 - 1 1 - - - - 1 - 1 2 LUEHSE Luehea semannii Triana & Planch. Tilliaceae ------1 - - MALMSP Malmea sp. Annonaceae ------1 - - MANIBI Manilkara bidentata (A. DC.) A. Chev. Sapotaceae - 2 - 1 2 - 1 - 1 - 4 - MAQUCO Maquira guianensis subsp. costaricana ( Standl. ) C.C.Berg Moraceae 2 - 1 ------MARAPA Maranthes panamensis (Standl.) Prance & F. White Chrysobalanaceae ------2 1 2 - 9 - MAR1LA Marila laxiflora Rusby Clusiaceae - 2 6 8 7 5 3 5 - - 8 10 MATAAP Matayba apetala Radlk. Sapindaceae 2 - - 1 - 1 2 - 5 - 1 - MICOAF Miconia affinis DC. Melastomataceae - 1 ------MICOLI Miconia ligulata Almeda Melastomataceae - - 3 3 5 - - 1 - - - - MICONE Miconia nervosa (Sm.) Triana Melastomataceae - - 1 ------MICOSI Miconia simplex Triana Melastomataceae - - - 5 1 1 1 2 1 - - 2 MICOSP Miconia sp. 1 Melastomataceae - - 4 ------MICO3 Miconia sp. 3 Melastomataceae - - 1 - 1 ------MOLLDA Mollinedia darienensis Standl. Monimiaceae - - - - - 2 3 - - - - - MORTAN Mortoniodendron anisophyllum (Standl.) Standl. & Steyerm. Tiliaceae - - 2 - 1 ------MOURMY Mouriri myrtilloides subsp. parvifolia (Benth.) Morley Melastomataceae - - 1 - - - - 1 1 - - - MYRCGA Myrcia gatunensis Standl. Myrtaceae ------1 - MYRCZE Myrcia zetekiana ( Standl. ) B.Holst Myrtaceae 1 - 1 - 2 - 1 - - 1 - - MYR22 Myrciaria sp. 2 Myrtaceae - - - - 1 ------NECTPU Nectandra purpurea (Ruiz & Pav.) Mez Lauraceae 3 - - 1 - - 1 - 1 - 1 - NEEAAM Neea amplifolia Donn. Sm. Nyctaginaceae 1 ------1 - - - OCOTDE Ocotea dendrodaphne Mez Lauraceae ------2 - 1 1 - 1 OCOTIR Ocotea insularis (Meisn.) Mez Lauraceae - - 2 1 1 ------1 OCOTPU Ocotea puberula (Rich.) Nees Lauraceae - - 1 ------OENOMA Oenocarpus mapora H. Karst. Arecaceae 1 7 - 3 2 5 2 2 2 - 2 2 OURACO Ouratea prominens Dwyer Ochnaceae - - 1 1 - - 1 2 - 2 - - OXANLO Oxandra longipetala R.E. Fr. Annonaceae ------1 - - - - OXANPA Oxandra panamensis R.E. Fr. Annonaceae ------7 - - PENTMA Pentagonia macrophylla Benth. Rubiaceae - - 1 1 - - - 1 2 4 - 1 PERAAR Pera arborea Mutis Euphorbiaceae - 1 ------PEREXA Perebea xanthochyma H. Karst. Moraceae 4 13 9 12 2 1 3 5 2 5 8 5 PICRLA Picramnia latifolia Tul. Picramniaceae ------4 - - PIPEA1 Piper arboreum subsp. arboreum Aubl. Piperaceae - - 1 ------PIPECU Piper colonense C. DC. Piperaceae 9 ------PIPECO Piper cordulatum C. DC. Piperaceae ------1 - - - - - PIPERE Piper reticulatum L. Piperaceae 3 ------PIPESP Piper sp. 1 Piperaceae 2 ------PIPE5 Piper sp. 5 Piperaceae - - 1 ------PIT1RU Pithecellobium sp. Fabaceae:Mimos. 1 ------POSOLA Posoqueria latifolia (Rudge) Roem. & Schult. Rubiaceae ------2 - - POULAR Poulsenia armata (Miq.) Standl. Moraceae - - 2 ------POURBI Pourouma bicolor Mart. Cecropiaceae 1 1 1 - - - - 1 1 - - - POUTRE Pouteria reticulata (Engl.) Eyma Sapotaceae - - 2 1 1 1 - 1 - 1 - - PRI2CO Prioria copaifera Griseb. Fabaceae:Caesal. ------1 - - IBISCA - Panama 69

PROTGL Protium glabrum (Rose) Engl. Burseraceae - - 2 - - - - - 1 - - - PROTPA Protium panamense (Rose) I.M. Johnst. Burseraceae 4 7 1 1 3 3 3 7 7 - 14 6 PROTTE Protium tenuifolium Engl. Burseraceae ------2 - - PSE2SP Pseudolmedia spuria (Sw.) Griseb. Moraceae 1 5 ------PSYCCH Psychotria chagrensis Standl. Rubiaceae 2 ------PSYCG3 Psychotria grandis Sw. Rubiaceae ------1 - - PSYCHO Psychotria horizontalis Sw. Rubiaceae 76 ------PSYCSU Psychotria suerrensis Donn. Sm. Rubiaceae - - 3 - 1 1 1 3 2 - - 1 QUIISC Quiina schippii Standl. Quiinaceae - 1 ------RANDAR Randia armata (Sw.) DC. Rubiaceae ------5 - - RINOSQ Rinorea squamata S.F. Blake Violaceae ------26 - - SAPISP Sapium sp. ‘broadleaf’ Euphorbiaceae 1 ------SIMAAM Simarouba amara Aubl. Simaroubaceae 1 2 1 - - - - - 1 - - - SLOAM1 Sloanea sp.aff.meianthera Donn. Sm. Elaeocarpaceae - - 2 1 - - - 2 1 - 1 1 SOCREX Socratea exorrhiza (Mart.) H. Wendl. Arecaceae - 7 3 3 2 5 8 7 9 - 5 6 SOROAF Sorocea affinis Hemsl. Moraceae 8 - 2 - 1 - - - 1 1 - - SPONMO Spondias mombin L. Anacardiaceae ------1 - - STEMGR Stemmadenia grandiflora (Jacq.) Miers Apocynaceae ------1 - SWARS2 Swartzia simplex var. continentalis Urb. Fabaceae:Papil. - 1 1 1 4 - 2 - 1 - 1 1 SYMPGL Symphonia globulifera L. f. Clusiaceae - - 1 - 1 1 1 1 1 - - - SYNEWA Synechanthus warscewiczianus H. Wendl. Arecaceae - - 3 3 5 - - 3 2 - 3 4 TAB2AR Tabernaemontana arborea Rose in Donn. Sm. Apocynaceae - - - - 1 ------TACHVE Tachigali versicolor Standl. & L.O. Williams Fabaceae:Caesal. 1 - 2 1 - - 1 1 2 - - 1 TALIPR Talisia princeps Oliv. Sapindaceae ------1 - - TAPIGU Tapirira guianensis Aubl. Anacardiaceae 2 1 5 - 1 4 3 2 - - 5 3 TERMAM Terminalia amazonia (J.F. Gmel.) Exell Combretaceae 3 ------TERNTE Ternstroemia tepezapote Schltdl. & Cham. Theaceae 1 ------TET4JO Tetrathylacium johansenii Poepp. in Poepp. & Endl. Flacourtiaceae ------1 - - THEOBE Theobroma bernoullii Pittier Sterculiaceae - 1 - 1 3 - - - 2 - 2 - TOCOPI Tocoyena pittieri (Standl.) Standl. Rubiaceae ------1 - TOVOLO Tovomita longifolia (Rich.) Hochr. Clusiaceae 3 3 1 6 6 2 3 11 8 - 12 - TOVOST Tovomita stylosa Hemsl. Clusiaceae - 1 3 8 8 5 4 2 3 - 8 1 TRATAS Trattinnickia aspera (Standl.) Swart Burseraceae - 2 ------TRI2PL Trichilia pleeana (A. Juss.) C. DC. Meliaceae ------1 - - TRI2TU Trichilia tuberculata (Triana & Planch.) C. DC. Meliaceae ------4 - - UNONPA Unonopsis panamensis R.E. Fr. Annonaceae - 2 5 3 5 2 9 6 2 - 2 3 VIROEL Virola elongata (Benth.) Warb. Myristicaceae - - 3 - 1 1 1 - - - - - VIROSP Virola multiflora (Standl.) A.C. Sm. Myristicaceae 5 1 - 1 - 1 5 - 1 1 - - VIROSE Virola sebifera Aubl. Myristicaceae 3 1 1 1 1 1 1 5 5 - 3 1 VIROSU Virola surinamensis (Rol. ex Rottb.) Warb. Myristicaceae 5 - 2 ------3 - - VISMBA Vismia baccifera (L.) Tr. & Pl. Clusiaceae - - - - - 1 - 1 - - - - VOCHFE Vochysia ferruginea Mart. Vochysiaceae - - - 1 - 1 ------XYL1MA Xylopia macrantha Triana & Planch. Annonaceae - - 2 2 1 9 5 2 3 - - 4 ZANTPR Zanthoxylum acuminatum subsp. juniperinum (Poepp.) Reynel Rutaceae 1 ------??? Unidentified trees 4 2 3 2 - 3 - 2 1 - - -